US3545610A - Photometric sorting apparatus - Google Patents

Description

United States Patent Leonard Kelly; James F. Hutter, Peterborough; George R. Mounce, Wlllowdale, Ontario, Canada lnventors Appl. No. 789,999

Filed Jan. 9, 1969 Patented Dec. 8, 1970 Assignee Sphere Investments Limited Nassau,

Priority Feb. 2, 1968 Great Britain No. 5527/68 PHOTOMETRIC SORTING APPARATUS 20 Claims, No Drawings U.S. Cl. 209/73, 209/1 11.7:356/73;209/75 Int. Cl B070 5/342 Field of Search 356/72, 73; 209/ll1.5, 111.6, 111.7, 111.8, 75;250/223 [56] References Cited UNITED STATES PATENTS 2,717,693 9/1955 Holmes 209/11 15 2,882,785 4/1959 Biesele,.lr. 356/73 3,011,634 12/1961 Hutter et a1 209/111.8X 3,025,961 3/1962 Yetterland 209/111.5X 3,052,353 9/1962 Pritchett 209/111.5 3,075,641 1/1963 Hutter et a1 209/11l.5

Primary Examiner-Allen N. Knowles Attorney-Weir, Marshall, MacRae & Lamb ABSTRACT: An apparatus for sorting irregularly shaped objects by making repeated scans across each object to detect light reflected from the objects surface. The apparatus derives from the reflected light at least 3 parameters and compares the parameters in a predetermined manner. An object is accepted or rejected according to the comparison.

1 PHOTOMETRIC SORTINGAIPARATUS BACKGROUND OF THE INVENTION This invention relates to apparatus for sorting ore and the like, and in particular it relates to apparatus which uses light reflected by .the pieces of ore or other objects to obtain a plurality of parameters on which the sorting is based.

While the apparatus may be used for sorting various kinds of irregularly shaped objects, it will be described hereinafter in connection with the sorting of ore.

For convenience, in the following description, the term pieces of rock will be considered to refer to pieces or fragments which may be undesirable or waste pieces, to pieces or fragments which may be desirable or valuable, or to pieces or fragments which contain both. waste and a valuable constituent. 1

One type of known apparatus for sorting ore and other objects by the use of light, illuminates the objects one at a time and collects the light reflected from the object. The light is then compared to a predetermined light value and if the light exceedsthis value the object is accepted. If the light does not exceed this value the object is rejected.

Another type of known apparatus for sorting ore and other irregularly shaped objects feeds the objects through a sorting zone in a random stream where the objects have any random orientation but do not rest one on another. That is, the objects are in a stream one layer thick thick. The stream of objects is illuminated and a light detector makes repeated scans across the field to detect the amount of light reflected by the surfaces of the objects in the path of the scan. The stream is divided, in effect, by a plurality of rejection devices in abutting relationship across the stream and located downstream of the light detector. That is, each rejection device may be said to define a channel extending along the stream. If the reflected light received by the light detector for an object moving along a particular imaginary channel is less than a predetermined value, the rejection device for that channel is actuated at a time when the object is passing it. Thus, an object may be accepted or rejected according to the reflecting quality of its surface.

It is relatively simple to sort objects such as pieces of rock when the valuable constituent is black and the waste white, or vice versa versa. The S( rting maybe readily done on the basis of the amount or the mplitude of the reflected light. However, the variety of cor ipositions of rock more is boundless. For example, a piece of rock might comprise a waste component of varying shades of grey and green, an undesired component in the form of lighter coloured flakes, and a valuable or desired component in the form of darker pebbles or small particles. It would be difficult to sort pieces of rock from an ore body of this type solely on the basis of the average amount of light reflected by each piece. Indeed, it would be difficult to sort on this basis pieces of rock taken from ore bodies in many parts of the world. It would be desirable to have an ore sorting apparatus which could be adapted to sort pieces or rock taken from any one of a large number of ore bodies having a variety of different ores.

SUMMARY OF THE INVENTION The present invention provides an apparatus which is versatile and can be adapted to sort rock from any one of a large number of ore bodies. The apparatus may be termed a photometric apparatus because it makes use of amounts of reflected light. However, it does not use only the amount of reflected light as the sole basis for making a sorting decision. Rather it derives a plurality of parameters from the reflected light and bases the sorting'decision on a combination of these parameters.

Thus, the present invention is for an apparatus for sorting irregularly shaped objects such as pieces of rock comprising means to move the objects through a sorting zone in the apparatus, means to make repeated scans across the sorting zone to detect-light reflected from discrete areas of each object as the scan traverses the object, means to derive a series of at least 3 parameters from the reflected light, means to select from the series of parameters a plurality of parameters for comparison, means to compare the plurality of parameters in a predetermined manner, and means responsive to the comparison to accept or reject a respective object.

It is because the apparatus makes use of a plurality of parameters that it is more versatile. The parameters may be used singly or in different combinations when one parameter may be given more weight than another may overrule another, depending on the particular objects being sorted.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a side view of apparatus according to the invention,

FIG, 2 is a partial end view of the apparatus shown in FIG.

FIG. 3 is a simplified view of 'a scanning apparatus suitable for the invention,

FIGS. 4 and 5 are block diagrams of circuitry used in the apparatus of the invention, and

FIG. 6 is a series of waveforms that occur in the circuitry and are useful in explaining the operation of the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The apparatus described is adapted to sort pieces of rock as the pieces move in a wide path random stream through the apparatus. The expression wide path as used herein is intended to mean a path of travel having sufficient width to permit a plurality of pieces of rock to move along the path beside one another. The term random stream as used herein is intended to mean that the pieces/of rock moving in a given direction have a haphazard alignment and spacing.

In the following description the words white, whiter and related words, and the words black, blacker and related words, are used. These words are used for convenience to denote relatively lighter and darker surfaces. In other words, a coloured surface may be referred to as having whiter areas and blacker areas, although the areas are not technically white or black. Similarly, if pieces of rock are scanned by a coloured light, for example a light from a laser source which has only a single wavelength in the red region, the reflected light received by the light detector or the signal from the light detector may be referred to as being whiter than a certain level or blacker than a certain level.

Thus, for example, in the apparatus of the invention the scan signal of reflected light is used to derivesignals representmg:

a. the variable amplitudeof the reflected light which is whiter than a predetermined value and which may be referred to as a white linear signal,

b. the amount of rock scanned where the reflected light indicates the rock is whiter than a predetermined value,.

which is the output of a passed through a squarer circuit and which may be referred to as a white squared signal,

c. the variable amplitude of the reflected light which is blacker than a predetermined value and which may be referred to as a black linear signal,

. the amount of rock scanned where the reflected light indicates the rock is blacker than a predetermined value, which is the output of c passed through a squarer circuit and which may be referred to as a black squared signal,

e. the size of the piece of rock as indicated by that scan, and

f. the number of changes from a predetermined light or white value to a predetermined black or dark value, and vice versa, where the rate of change exceeds a predetermined value, and which may be referred to as the number of counts.

- Referring now to FIG. 1, a hopper 10 is shown holding a quantity of pieces or fragments 11 of rock. The pieces of rock move downwards ,under the influence of gravity and are deposited on a vibrating table 12 which is suspended by springs 14 or by any other mounting means which will permit vibration. The table 12 is driven or vibrated by a motor 15. Vibrating table feeders are well-known in the art. The pieces of rock move along the surface of table 12 and are discharged at the end onto another vibrating table 16 suspended by springs 17 and driven by motor 18. The vibrating table 16 may have a narrow slotted portion or a screened section to permit undersize pieces and small particles to fall through to a disposal (not shown). Pieces of rock are received at one end of table 16, move along the surface, and are discharged onto a slide plate 20.

The speeds of vibration of tables 12 and 16 are preferably each independently adjustable. The adjustment of the two speeds permits a closely packed but single layer of pieces to be formed towards the discharge end of table 16. This will give an optimum rate of sorting.

The pieces of rock fall onto slide plate 20 past a chain curtain 21 which helps prevent rocks from rolling as they move onto slide plate 20. The pieces of rock accelerate as they slide down plate 20 and are discharged onto a moving belt 22. By way of example only, a typical speed of movement of the pieces of rock at the discharge end of table 16 might be of the order of 40 feet per minute, while the belt speed might be of the order of 200 feet per minute. The slide plate 20 serves to accelerate the pieces of rock to a speed which is closer to belt speed, and the end of slide plate 20 is curved so that it is parallel to belt 22 to deliver the pieces of rock gently onto belt 22 to minimize wear and to minimize the tendency for the rocks to roll. Two skirts 23 on either side of belt 22 limit the lateral placement of rock pieces on the belt.

The endless belt 22 is supported by idlers 24 between the head roller 25 and a drive roller 26. Roller 26 is driven by a motor (not shown). A spray 27 is positioned below belt 22 just preceding a driven rotating'brush 28 to wash and clean the belt.

As the rocks slide down plate 20 and are deposited on belt 22 they are accelerated and this increases the spacing between pieces of rock in the direction of movement. This increase in spacing is desirable for the sorting in that there is a detectable space between one piece of rock and the next.

The pieces of rock move along on belt 22 past a scanning device 30 and are discharged in a free fall trajectory at head roller 25 as shown in FIGS. 1 and 2. The rocks pass a rejection mechanism 31 which may comprise a number of adjacent-air blast nozzles 32 arranged side by side extending across the width of the stream of rocks. A guard plate 37 immediately above nozzles 32 is provided to protect the nozzles from accidental damage. The guard plate 37 is normally not touched by the pieces of rock as they pass in free fall and provides protection against abnormal rock position. There are 10 nozzles shown, but it will be apparent that any number suitable to the type bf rock, sorting rate, etc. could be used. The rejection mechanism includes a control valve for each nozzle 32 and an air source, and, when a piece of.rock is to be rejected, the necessary valve is opened to direct an air blast on the piece of rock and deflect it to one side of splitter plate 33 to be carried away by belt 34. Those pieces of rock which are of sufficient value are permitted to fall onto belt 35 for subsequent processing.

The rejection mechanism 31 and the zone in which scanning takes (i.e.) the zone where the rocks pass scanning device 30) are separated by a distance of several feet. This separation is desirable because the air blast rejection generates some splash and mist which could interfere with the optical scanning of device 30. It may also be desirable to separate the scanning zone from the zone where rejection takes place by a baffle 36 or alternately to enclose, an area with a bafile on either side of the scanning device 30 and provide a downward flow of clean air in the enclosed area to keep dust and mist from the optical path.

It may be desirable to have belt 22 of a particular colour to aid in sorting. lfa dark rock is being sorted it may be desirable to use a white belt to provide a contrasting background for the scanning device, and ifa lighter rock is being sorted it may be desirable to use a black belt.

The scanning device 30 may be of any type that will provide a continuously repeated optical scan across the stream of pieces of rock on belt 22.Many such scanning devices are known in the art. Some of these 'devices make use of polarized light to avoid problems with specular reflection. When scanning objects such as pieces of rock, the optical scanning detector could receive specular light reflected from small, highly polished, reflecting surfaces and diffuse light reflected generally from other surfaces. Specular light reflected from a given area may be many times the value of diffuse light reflected from an area the same size. It is the diffuse light that gives a more precise indication of surface quality, and this could be masked by specular reflection. In order to overcome this the objects are illuminated by light polarized in a first direction, and the reflected light going to the scanning detector is passed through a polarizing filter polarized in a second direction at right angles to the first direction. The apparatus of the present invention makes use of polarized light in this manner.

In the optical scanning of objects the objects may be placed in an illuminated area and the light detector caused to scan across the objects to detect the diffuse reflected light. Such a scanning system is described, for example, in British Specification No. 986,177. It is, however, preferable to have the light source scanning in conjunction with the detector as this enables a greater illumination to be achieved in the area of importance, that is, in the area being scanned by the light detector. The apparatus of this invention preferably uses such a scanning means, and a suitable one is shown in FIG. 3.

Referring now to FIG. 3, there is shown a light source 40 which is preferably a laser. A laser provides a convenient and well defined spot oflight in the scanning zone, and in addition the light is polarized. The laser beam from source 40 is directed towards a rotating S-sided mirror drum 41 so that the rotation of drum 41 causes the light beam to travel or scan across the width of the belt. FIG. 3 is not drawn to scale and is intended only to indicate one example of a suitable scanning arrangement. The belt is normally a distance from the rotating drum which is many times the drum diameter. A light detector 42, which may be any suitable photo detector, is positioned to receive light reflected from the belt. The light detector 42 includes a light polarizing filter to screen out specular reflection as was previously discussed.

It will be seen that the light detector 42, in effect, scans across the belt in conjunction with the scanning of the light source 40. That is, they are both directed by the mirror drum 41 to the same spot on the belt carrying the pieces of rock, and as the drum 41 rotates this spot moves across the belt. By having both the light source 40 and the light detector 42 scan in conjunction with one another it is possible to reduce the unwanted or. random light to a minimum.

FIG. 3 shows mirror drum 41 in two positions as an aid to understanding. One position is shown in solid lines and one in broken lines. The light path is indicated for each position. It is believed that the operation will be clear.

It is, of course, necessary to know when the scan begins at one side of the belt and when it ends at the other side. This may be done in a number of ways, such as, for example, by having a number of slugs of magnetic material spaced appropriately on the sides of mirror drum 41. Alternately, the arrangement shown in FIG. 3 may be used where a photodiode 43 is placed at the beginning of the scan and a photodiode 44 is placed at the end. The light from light source 40 is reflected by mirror drum 41 onto photodiode 43 and onto the belt at one side. As the drum 41 turns the light sweeps across the belt, and as it reaches the other side it is directed onto photodiode 44. Thus, the outputs from photodiodes 43 and 44 may be used to define the time of the scan, and if necessary, this may be broken down electronically into any number of parts.

A white standard 45 is placed in the scanning path just to one side of the limit of the belt scan. The purpose of this is to provide a standard or reference level to stabilize the'output from the light detector 42.

FIGS. 4 and 5 toe together show the simplified block diagram of the circuitry used in the invention, and FIG. 6 shows various waveforms in different parts of the circuitry. The description will be better understood with reference to FIGS. 4, 5 and 6 in conjunction.

In FIG. 4, there is shown light detector 42 which receives light reflected from the scanning mirror. A polarizing filter is incorporated on the detector so the light passes through it before reaching the light sensitive element. The detector 42 provides an electrical output representing the diffuse light reflected by pieces of rock and the supporting belt in the sorting zone and also diffuse light reflected by the white standard reflector. The output from light detector 42 is applied to a DC amplifier 50. Also in FIG-.-4 is shown schematically, the two photodiodes 43 and 44 which receive light at the beginning and end of each scan. These photodiodes provide outputs to a timing register 51 which, in turn, provides two output signals. Timing register 51 provides pulses representing the timing output signals from the two photodiodes to a gate timing circuit 52, a black clamp timing circuit 53, and a peak white timing circuit 54. Atypical waveform is show shown by way of example at a in FIG. 6, Timing register 51 also provides a number of channel outputs. That is, timing register 51 divides the time between the start of the scan and the end of the scan into a number of equal portions according to the number of air blast nozzles. In the embodiment described there are ten such nozzles, and timing register 51 therefore divides the time of scan into portions. It provides ten outputs each of which comprises a pulse lasting for a particular one-tenth of the scan. These outputs, in effect, divide the scan into ten channels. For simplicity of drawing only one such output is shown at conductor 55. For example, a waveform which is typical of one of the channel dividing outputs is shown at b in FIG. 6. The waveform at b represents the time period the scan is in channel 3, and this will be discussed subsequently in cone connection with FIG. 5.

The gate timing circuit 52 receives the two timing pulses from timing register 51 and provides a gating pulse representing the actual scan time.'This is used in circuits to be described later to gate the circuits on only for that period of time when the scan is passing across the belt, i.e. across the actual sorting zone. A typical waveform representing the output of gate timing circuit 52 is shown at c in FIG. 6.

The black clamp timing'circuit 53 and the peak white timing circuit 54 are used to stabilize the output of DC amplifier 50. It is known that DC amplifiers tend to drift and that some form of stabilization is desirable. The type of stabilization indicated in FIG. 4 is known and is described, for example, in US. Pat. No. 3,097,744 to J. F. Hutter et al., and willbe describedvery briefly herein.

Considering first the peak white timing circuit 54, this provides an output such as is shown at d in FIG. 6 and this output is used to gate on an automatic gain control circuit 56 during the time the lig ht detector is receiving light reflected from the white standard 45. Thus, if the light level of the scanning beam should increase for any reason, the detector 42 will receive more reflected light and particularly more light reflected from the white standard 45, the automatic gain control circuit 56 will detect the increase in the light reflected from standard 45 and will decrease the sensitivity of detector 42 correspondingly.

Considering now the black clamp timing circuit 53, this provides an output such as is shown by the waveform at e in FIG. 6. It will be noted that the waveform is in the form of a square pulse which occurs after the scan, that is which occurs when the light detector 42 should be receiving no reflected light. During the time of this pulse the output from amplifier 50 should be at a zero level or a baseline level. Thus, the output of black clamp timing circuit 53 is applied to DC amplifier 50 and gates on a control incorporated in the amplifier to ensure that the output of amplifier 50 is at a baseline level during the time of the gate.

The output from DC amplifier 50 is the main output signal or the video output signal and is a signal representing the reflected light received. It is, of course, dependent on the pieces of rock scanned, but for the purpose of this description the main video output for one scan may be represented by the waveform at f in FIG. 6. The main video output is applied to several circuits. One of these is'an area or size discriminator and squarer circuit 57 which also receives a gating signal from gate timing circuit 52. The gating signal ensures that the area discriminator and squarer 57 can only produce an output when the scanning beam is sweeping the sorting zone. The discriminator is set at a level quite close to the signal level which is produced when the scan is on the white supporting belt, that is, it is set at a level just darker than the belt as is indicated at the right of the waveform at f in FIG. 6. The discriminator output is squared, that is, formed into a square-shaped pulse, and the squared output is provided on conductor 58. This output would be as shown at g in FIG. 6 and represents the time periods when the scan is on a piece of rock, or in other words, it represents the size of the piece of rock as seen by that scan.

The main video output and the output from the area discriminator and squarerare each applied to four circuits. These circuits are all discriminator type circuits whose levels of discrimination may be individually varied. They are gated on by the output from area discriminator and squarer 57. The four circuits are: a black discriminator 60, a white discriminator 61, a white swing discriminator and squarer 62, and a black swing discriminator and squarer 63.

The black discriminator 60 produces an output whenever the detected signal is blacker than a predetermined level. If we assume the black discriminator 60 has a discriminating level shown in FIG. 6 at the right side offas black swing, then for the waveform at f the output of black discriminator 60 would be substantially as shown at h in FIG. 6. The output is a linear function of the blackness of the rock and may be referred to as the black linear signal. The output of discriminator 60 is available on conductor 64 and is also applied to a squarer 65 which produces constant amplitude pulses corresponding to widths of portions of rock being scanned which are darker than the discriminator level. The output from squarer 65, for the waveform shown in FIG. 6, f, would be substantially that shown at i in FIG. 6. This output is available on conductor 66.

Similarly the white discriminator 61 produces an output whenever the detected light has a value above a predetermined level, that is whenever the signal is whiter than a predetermined level. It should be remembered that the discriminator 61 is gated on only when the scan is actually traversing a piece of rock. The output from discriminator 61 is a linear function of the whiteness of the rock and may be referred to as the white linear signal. This output is available on conductor 67 and is also applied to a squarer 68 which produces constant amplitude pulses corresponding to widths of portions of rock being scanned which are whiter than the discriminator level. The output from squarer 68 is available on conductor 70.

Assuming the discriminator 61 is set at a level shown as white swing in FIG. 6, f, then the outputs from discriminator 61 and squarer 68 would be substantially as shown at j and k respectively.

The white swing discriminator and squarer 62 operates in the same manner as the combination of white discriminator 61 with squarer 68. That is, the output from white swing discriminator and squarer 62 would be of the same type as that on conductor 70. In order to keep to a minimum the number of discriminator levels shown in FIG. 6, f, it has been assumed that the discriminator 61 and the discriminator portion of white swing discriminator and squarer 62 are set at the same level, i.e. at the white swing level of FIG. 6,f. Thus, in this particular instance, the output from white swing discriminator and squarer 62, for the same waveform of FIG. 6,fwould be substantially that on conductor 70 which is indicated in FIG. 6, k. Similarly, in this particular instance, the output from black swing discriminator and squarer 63 would be substan tially that on conductor 66 which is indicated in FIG. 6, i.

The outputs of swing discriminators and squarers 62 and 63 are fed to variable stretch circuits 71 and 72 respectively. These stretch circuits 71 and 72 extend the pulse duration by an amount which can be set into the circuits. The outputs of stretch circuits 71 and 72 are applied as inputs to AND circuit 73. The AND circuit 73 provides an output only when there is a pulse present at both inputs. In other words, AND circuit 73 provides an output only when the main video signal swings from a black level to a white level within the stretch time, or from the same white level to the same black level within the stretch time. The rate of change from black to white, or vice versa, which will produce an output from AND circuit 73, may be altered by altering the stretch time. The output from AND circuit 73 is therefore a number of pulses representing a count" the times the video swings from black to white and vice versa at a rate greater than a predetermined rate.

It will be apparent that, for example, when a black piece of rock is on a white belt, there will be a count at the beginning and at the end of the scan across the rock. That is, as the scan moves across the belt it is on white or it sees" white, and as it comes to the edge of the rock it is on black or it sees" black. As the scan traverses the rock the video signal changes from belt white to the black of the rock at one edge, and from the black of the rock to belt white at the other edge. These may both register as counts but they do not represent changes in the reflectivity on the surface of the piece of rock. It is, of course, desirable to eliminate these spurious counts.

The count as the scan swings from belt white to black presents little problem. The swing discriminators and-squarers 62 and 63 are gated by area discriminator and squarer 57. Because there is a small delay inherent in the gate signal from discriminator and squarer 57, the white swing discriminator and squarer 62 is not gated on at the instant the scan passes from the belt onto the piece of rock. Thus, there will be no count as the result of the scan passing from belt white onto a black rock. However, the delay in the gate signal is a disadvantage as the the scan passes from the black piece of rock to belt white. There will be a definite interval between the time when the video signal is above the white swing level but the signal from the area discriminator 57 has not yet cut off the discriminator and squarers 62 and 63. Thus, as the scan passes from the rock onto the white belt, and the AND gate 73 will receive a stretched black pulse (from the rock) and a white pulse (as the scan just passes onto the white belt). This will produce an output pulse or a count from AND gate 73. The output from AND gate 73 is fed to a gated pulse delay circuit 74, which is gated by the area signal from the area discriminator and squarer 57. The delay in the gated pulse delay circuit 74 is adjustable and is set so that the count caused by the scan leaving a dark rock and passing onto belt white occurs just after the end of the gate signal from discriminator and squarer 57. This serves to inhibit the count generated as the scan leaves the piece of rock. The output of gated pulse delay circuit 74 thus represents the number of counts as previously described and is available on conductor 75.

For convenience in reading from FIG. 4 onto FIG. 5, conductors 55, 75, *0, 67, 66, 64 and 58 have been given terminal designations A nrough G respectively.

Referring now to FIG. 5, it will be seen that the signals at the terminals designated A through G are as follows:

A a particular channel gating signal B the signal representing the number ofcounts C the white squared signal D the white linear signal E the black squared signal F the black linear signal G the sizesignal In FIG. 5 there are four switches 76, 77, 78 and 79 each of which has six positions. The contacts for the six positions in each of the switches 76-79 are connected to one of the terminals designated BG. That is, each switch may be positioned to select one of the six signals available at BG. Thus, the movable contact of each of the switches BG will have one of the signals on it.

So far the circuitry described has been to provide signals representing several parameters. The circuitry to be described below is provided for each channelfOnly the circuitry for one channel is shown, and this may for example be for channel 3 as indicated by the waveform b in FIG. 6 which is the signal at A. A line in FIG. 5 indicates where the circuitry for the individual channels begins. I v

There are four variable gated integrators 81 82, 83 and 84 each of which has three inputs. One of the inputs in each integrator is connected to A to receive the gating signal for that channel. Another input of integrators 8I84 is respectively connected to the movable contact of switches 76-79. The third input is connected to a reset circuit 86. The reset circuit 86 is connected to A to receive a channel signal and to G to receive a size signal. The operation of the circuits is quite straightforward. Each of the integrators 81'84 is gated on while the scan is crossing the channel associated with those integrators, and if there are signals present on the movable contact of the switches 7679 connected to that integrator, the integrator will sum the signals. When the integrator receives a reset signal from reset circuit 86, it provides an output proportional to the integratedsignals. The outputs for integrators 81- -84 are on conductors 87-90 respectively. The reset circuit 86 provides a reset signal when the scan does not traverse a rock for that channel. That is, reset circuit 86 provides an output or reset signal when it does not receive a size signal from G during the time when it is gated on by the channel signal from A.

The conductors 87 and 88 are connected to a comparator 92 to provide inputs therefor, and conductors 89 and 90 are connected to a comparator 93 to provide inputs therefor. The outputs of comparators 92 and 93 are connected to a comparator 94 to provide inputs therefor. Each of the comparators 92, 93 and 94 may be set to provide an output a when one input is greater than the other, b when the ratio of one to the other exceeds or is less than a predetermined value, or c when the sum or difference is greater or less than a predetermined value.

The output from comparator 94 is connected to a timing and air blast control 95. When the various parameters are such that a piece of rock is to be deflected by an air blast, the signal is received from comparator 94 and the timing and air blast control circuit 95 provides a time delay sufficient for the piece of rock to reach the air blast nozzles 32 (FIG. 3) and it actuates the air blast control for the nozzle in that channel to direct a blast of air at the piece of rock.

It will be apparent that more accurate timing could be obtained by providing an additional circuit which recognizes the leading edge ofa piece of rock in a particular channel by the initiation of the first size signal, and which receives from reset circuit 86 a signal defining the trailing edge. If the scanning spot is well defined and the scan rapid, it is possible to determine quite accurately the duration of the air blast that is required.

It will also be apparent that the 6 parameters could be use, if desired, to feed various arrangements of comparators to provide a sorting decision.

In order to provide a better understanding of the operation of the circuitry, we will consider briefly the apparatus as it might be set up for sorting one particular ore. As an example, one of the main gold bearing ores in South Africa presents some difficulty in machine sorting where the sorting is based only on the average amount of light reflected. In one of the main gold bearing ores, gold is associated with certain conglomerate beds commonly referred to as reef. This term reef describes rock having a random scatter of light coloured, roughly spherical, quartz pebbles of varying sizes embedded in a matrix of darker material. This matrix generally comprises recrystallized quartz grains, pyrite and other sulphides with other secondary constituents.

The reefs are generally narrower than minimum mining width, so a proportion of surrounding rock must be taken with the reed reef. This proportion may be 50 percent or even higher. The associated waste rock is generally of two types as follows:

l. a dark tulfaceous lava which may be referred t to as tuff," and 2. a greenish-grayish quartzite.

Speaking very generally, it is correct to say that reef has light coloured pebbles in a dark matrix, the quartzite has a fairly uniform gray or greenish colour, and the tuff has a fairly uniform blackish colour. There are, however, different shades of black, white and grey in all the rock types. in fact there is a considerable overlap in response. Light areas of the tuff and quartzite are as light as some of the pebbles, and some of the dark areas of the quartzite are as dark as some of the reef matrix. This makes it difficult to sort the ore from the waste on the basis of the amount of light reflected.

It is reasonably easy for a trained person to visually identify a piece of rock of this type as ore or waste. The eye takes in the overall appearance of the rock, and particularly the characteristic roundish section of the broken pebbles. However a sorting machine such as is known in the prior art detects the amplitude of the reflected light, and because of the variations, a consistent and satisfactory sorting decision cannot be made.

It has been found necessary to scan across a piece of rock several times and to accumulate the information received from the successive scans in order to make a satisfactory sorting decision. it has also been found necessary to derive from the light recieved by the light detector a plurality of parameters on which to base the sorting decision.

A study of the type of ore being used in this example indicated that a significant factor was the rapidity of change or swing from white to black, and black to white, as the scan traversed the junction of a dark matrix and a white pebble. Similar swings do occur in the quartzite rock but they are less frequent and the swing is often not as rapid. Therefore one significant sorting basis could be the number of such swings per unit area, that is the counts per unit area. This would make a reasonable separation between reef and quartzite.

The discrimination between reef and tuff requires an additional factor. The tuff pieces frequently have small white areas which give a signal of counts per unit area greater than those derived from reef pieces. However, the tuff pieces normally have a greater percentage area of black than does the reef. A reasonable separation of reef from tuff might be made by requiring an increased number'of counts per unit area if the black area per unit area exceeds a certain level.

Thus the signals required in the sorting decision would be signals proportional to l. the number of counts 2. the area of the piece of rock 3. the black area of the piece of rock.

From these three signals can be derived signals proportional to a. number of counts per area of rock b. black area per area of rock. I

These last signals arethe signals on which the sorting decision for this particular ro'ck type is to be based.

To adjust the apparatus for this the switch 76 is set to the first position to make a connection-with B to select the signal proportional to the number of counts; the switch 77 is set to the sixth position to make a connection with G to select the signal proportional to area; the switch 78 is set to the fourth position to make a connection with E to select the signal proportional to the black area; the switch 79 is set to the sixth position to make a connection with G 'to select the signal proportional to rock area. The four signals selected are integrated as has been described. The integrated signals proportional to counts and to rock area are applied to comparator 92 which provides an output proportional to counts per unit area provided that the counts per unit area are greater than a predetermined minimum. if the minimum is not exceeded a reject signal or a signal of a low reference value is provided as an output. if this minimum is exceeded the output signal proportional to counts per unit area is applied to comparator 94, and this is a tentative accept signal. a

The integrated signals proportional to the black area and to the rock area are applied to comparator 93. if the ratio of 'black area to total rock area is less than a predetermined value a no reject signal or a signal of a low reference value is provided as an output. if the ratio of black area to total rock area is greater than this value a tentative reject signal comprising the ratio of black area to rock area is provided. This output is applied to comparator 94.

Comparator 94 receives the two signals as described. if the ratio of these signals is greater than a predetermined value, the piece of rock should be accepted.lfthe ratio of these signals is less than this predetermined value the piece of rock should be rejected, and an appropriate output signal is provided to cause rejection. This will perhaps be made clearer by reference to the following table:

COMPARATOR 92 COMPARATOR 93 Black area Accept: Accept: Low ref. signal.

Total area Black area Black area Black area and total Reject: y Tentative reject:-- area. Total area Total area COMPARATOR 94 Flam comparators 92 and Reject: z

Sig. from comp. 92

Accept: z Accept: No signal.

Sig. from comp. 93

Sig. from comp. 92 Reject: Signal to rejection mechanism.

Sig. from comp. 93

The symbols x. y and z in the above table are predetermined levels chosen experimentally for the particular rock.

it will be seen that comparator 94 could receive four possible inputs as follows:

PAT. No. 3,545,610

it will be apparent that in l above the piece of rock will be accepted as the ratio will be high because of the low reference signal from comparator 93. In (2) above it will depend on the actual ratio whether or not the piece of rock is accepted or rejected. ln (3) above, there will be two low reference signals, and in this instance these signals are selected so that the signal from comparator 92 is overriding; that is, the signals are selected so that the ratio required by comparator 94 for acceptance will not be met. In (4) above, there will be a reject output because the ratio will be small due to the low reference signal from comparator 92.

The preceding example will help to show the versatility of the apparatus. The apparatus is able to sort a very great number of types of ore from various parts of the world. It is necessary only to study the ore, select suitable parameters for that ore, and decide on the overriding factors where necessary.

It will, of course, be apparent that the apparatus is able to sort irregularly shaped objects other than pieces of rock by scanning the objects and developing from the reflected scan a plurality of parameters on which the sorting of the objects may be based.

We claim:

1. Apparatus for sorting irregularly shaped objects comprismg:

means to move the objects through a sorting zone in the apparatus;

scanning means to make repeated scans across the sorting zone to detect light reflected from discrete areas of each object as the scan traverses the object;

means to derive a series of at least three parameters from the reflected light;

selecting means to select from said series of parameters a plurality of parameters for comparison;

comparator'means to compare the selected parameters in a predetermined manner; and

rejection means responsive to the comparison to reject an object.

2. Apparatus as defined in claim 1 and further including integrator means for each selected parameter to integrate the value of the respective selected parameter from successive scans across an object or across a predetermined portion of an object.

3. Apparatus as defined in claim 1 in which said plurality of parameters is at least 3 parameters.

4. Apparatus for sorting pieces of rock having ore and waste comprising:

means to move pieces of rock through a sorting zone in the apparatus;

scanning means to make repeated scans across the sorting zone including a light detector to detect light reflected from discrete areas of each piece of rock as the scan traversesits surface and to provide an electrical output representingreflected light;

circuit means receiving said electrical output and deriving therefroma series ofat least 3 parameters;

rality of parameters in a predetermined manner and to provide a sorting signal based on the comparison; and rejection means connected with said comparator means and responsive to said sorting signal of a predetermined value to reject a respective piece of rock. 5. Apparatus as defined in claim 4 in which said selecting means selects at least 3 parameters as said plurality of parameters for comparison.

6. Apparatus as defined in claim 4 in which said selecting,

means selects as said plurality of parameters for comparison at least three of the following parameters:

a. a white linear signal which is a variable amplitude signal representing reflected light which is whiter than a predetermined level;

b. a white squared signal which is of substantially constant amplitude and of a pulse duration corresponding to the time the reflected light is whiter than a predetermined level;

0. a black linear signal which is a variable amplitude signal representing reflected light which is blacker than a predetermined level;

d. a black squared signal which is of substantially constant amplitude andof a pulse duration corresponding to the time the reflected light is blacker than a predetermined level;

e. a size signal representing the size where traversed by the scan; and

f. a signal representing the number of counts where a count is a change from a predetermined white level to a predetermined black level, and vice versa, with the rate of change in excess of a predetermined rate of change.

7. Apparatus as defined in claim 6 and further including integrator means for each selected parameter connected between said selecting means and said comparator means to integrate the value. of the respective selected parameter from successive scans across a piece of rock or across a predeter- I mined portion thereof.

8. Apparatus as defined in claim 7 in which said comparator means comprises a plurality of comparators each having :two

inputs and an output, one of said comparators being adapted to provide at its output said sorting signal and being adapted to receive at at least one of its inputs the output from another comparator, said parameters constituting at least one of the inputs to at least two comparators.

9. Apparatus as defined in claim 7 in which said selecting means selects as said plurality of parameters for comparison a first, second and third parameter, and-in which said comparator means comprises a first, second and third comparator each having two inputs and an output,

said first comparator being connected to said selecting means to receive at its two inputs said first and second input parameter signals and to provide at its output a first output signal which is related to a comparison of the two input signals,

said second comparator being connected to said selecting means to receive at its two inputs said second and third input parameter signals and .to provide at its output a second output signal which is related to a comparison of the two input signals, and

said third comparator having its two inputs connected to the output of said first and second one of said comparators to receive said first and second output signal and to provide at its output said sorting signal which is related to a comparison of said first and second output signals.

10. Apparatus as defined in claim 8 in which at least one comparator is arranged to provide an output signal indicative of asignal at one of its inputs being greater than a signal at the other of its inputs.

11. Apparatus as defined in claim 8 in which at least one comparator is arranged to provide an output signal indicative of the ratio of the signal at one of its inputs to the signal at the other of its inputs lying outside a predetermined range of values.

12. Apparatus as defined in claim 8 in which at least one comparator is arranged to provide an output signal indicative of the sum or difference of the signals at its inputs lying outside a predetermined range of values.

13. Apparatus for sorting pieces of rock having waste and a valuable constituent, comprising:

a wide path handling means to move pieces of rock in awide path random stream through a sorting zone along a predetermined path; I

said path having a plurality of imaginary parallel channels;

scanning means including a light detector to make repeated scans across said path in said sorting zone to detect light reflected from discrete areas of each piece of rock as the scan traverses its surface and to provide a main video output signal representing the reflected light;

electronic means associated with said scanning means to provide a timing signal indicative of the scan passing the boundaries of said imaginary channels;

circuit means connected with said scanning means to receive said main video output signal and to derive therefrom a series of at least 3 parameters;

selecting means connected with said circuit means to select from said series of parameters a plurality of parameters for comparison;

comparator means connected with said selecting means and said electronic means to receive from said selecting means said plurality of parameters, to compare said plurality of parameters in a predetermined manner and to provide a sorting signal based on said comparison and related to the position of a piece rock with respect to said imaginary channels;

a plurality of rejection devices, one for each of said imaginary channels, in side by side abutting relationship corresponding to said channels and extending across the width of said path;

said rejection devices being positioned so that pieces of rock pass ,said rejection devices after passing said scanning means; and

said rejection devices being responsive to a sorting signal of a predetermined value for a respective channel to reject a piece of rock from which said sorting signal was developed.

14. Apparatus as defined in claim 13 in which said selecting means selects as said plurality of parameters for comparison at least three of the following parameters:

a. a white linear signal which is a variable amplitude signal representing reflected light which is whiter than a predetermined level;

b. a white squared signal which is of substantially constant amplitude and of a pulse duration corresponding to the time the reflected light is whiter than a predetermined level;

c. a black linear signal which is a variable amplitude signal representing reflected light which is blacker than a predetermined level;

d. a black squared signal which is of substantially constant amplitude and of a pulse duration corresponding to the time the reflected light is blacker than a predetermined level;

e. a size signal representing the size where traversed by the scan; and

f. a signal representing the number of counts where a count is a change from a predetermined white level to a predetermined black level, and vice versa, with the rate of change in excess of a predetermined rate of change.

lSQApparatus as definedin claim 14 and further including integrative means for each selected parameter connected to said selecting means to integrate the value of the respected selected parameter from a successive means across a piece of rock or predetermined portion thereof.

16. Apparatus as defined in claim 14 in which said wide path handling means comprises:

a hopper;

at least one vibrating table for receiving pieces of rock from said hopper and delivering from the end thereof a wide path random stream onto the upper surface of an inclined slide plate; and

a moving belt arranged to receive the wide path random stream of pieces of rock from the slide plate, carry said pieces of rock past said scanning means and discharge said pieces of rock past said scanning means and discharge said pieces of rock past said rejection devices.

17. Apparatus as defined in claim 14 in which said scanning means comprises a laser'light source directed onto a rotating, multi surfaced mirror to cause a spot of light to traverse the sorting zone, said light detector being arranged to receive diffuse light from said mirror reflected onto said mirror from pieces of rock in said spot of light.

18. Apparatus as defined in claim 14 in which said selecting means is switch means operable manually to select desired ones of said series of parameters.

19. Apparatus as defined in claim 14 in which said rejection devices each comprise an air blast nozzle and a source of air under pressure connected to said nozzle, and an air blast control valve to control the flow of air to said nozzle according to said sorting signal.

20. A method for sorting irregularly shaped objects comprising the steps of:

moving the objects through a sorting zone;

making repeated scans across the surface of each object to detect light reflected from its surface;

providing from said detected light an electrical output signal varying in accordance with the detected light;

deriving a series of at least 3 parameters from said electrical output signal; selecting from said series of parameters a plurality of parameters and integrating the value of each selected parameter for a particular object or portion thereof; comparing the integrated values of the selected parameters in a predetermined manner;

providing a decision signal based on the comparison representing a sorting decision; and

directing each object along one of two paths in accordance with the decision signal.

Images