WO2015005758A1 - Single-camera system for estimating the speed of vehicles by video processing with multi-camera calibration by stereoscopic effect - Google Patents
Single-camera system for estimating the speed of vehicles by video processing with multi-camera calibration by stereoscopic effect Download PDFInfo
- Publication number
- WO2015005758A1 WO2015005758A1 PCT/MA2014/000016 MA2014000016W WO2015005758A1 WO 2015005758 A1 WO2015005758 A1 WO 2015005758A1 MA 2014000016 W MA2014000016 W MA 2014000016W WO 2015005758 A1 WO2015005758 A1 WO 2015005758A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- camera
- cameras
- calibration
- speed
- processing unit
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 14
- 230000000694 effects Effects 0.000 title claims description 8
- 238000001514 detection method Methods 0.000 claims abstract description 4
- 238000004364 calculation method Methods 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000003287 optical effect Effects 0.000 claims description 4
- 238000011084 recovery Methods 0.000 claims 1
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000000034 method Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 244000045947 parasite Species 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S11/00—Systems for determining distance or velocity not using reflection or reradiation
- G01S11/12—Systems for determining distance or velocity not using reflection or reradiation using electromagnetic waves other than radio waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P21/00—Testing or calibrating of apparatus or devices covered by the preceding groups
- G01P21/02—Testing or calibrating of apparatus or devices covered by the preceding groups of speedometers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P3/00—Measuring linear or angular speed; Measuring differences of linear or angular speeds
- G01P3/36—Devices characterised by the use of optical means, e.g. using infrared, visible, or ultraviolet light
- G01P3/38—Devices characterised by the use of optical means, e.g. using infrared, visible, or ultraviolet light using photographic means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/254—Analysis of motion involving subtraction of images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
- G06T2207/10021—Stereoscopic video; Stereoscopic image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30236—Traffic on road, railway or crossing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
- G08G1/054—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed photographing overspeeding vehicles
Definitions
- This invention relates to video processing-based passive velocity estimation systems (speedmeters), which are used in particular for detecting violations of the highway code (speeding).
- the righting operation makes it possible to associate a metric distance with the pixels of the plane of the road.
- This operation requires to be made to have the metric coordinates of at least four points on the vehicle movement plan to deduce a grinding matrix. This is generally obtained either through a manual calibration (by an operator who determines the actual distance on the road between certain points), or by positioning on the pavement markings on the ground of which we know a priori the coordinates and which are easily identifiable on the images.
- Multi-camera systems are usually based on the stereoscopic effect that calculates the distance between the target and the axis of the cameras-Stereo by measuring the shift of the object on the left and right image taken at the same time (EP Patent No. 1067388 A1), these systems make it possible to make speed estimates without the need to obtain prior metric information on the road map.
- Our invention has the advantage of allowing the calibration of the single-camera system without calling on an outside operator or a ground marking. It can be carried out periodically (for example every day) which makes it possible to ensure that the system is permanently calibrated and that variations in positioning of the camera have no impact on the accuracy of the estimates of the system. in addition, it has the advantage that a stereoscopic multi-camera system can only occasionally use the second camera for calibration. The second camera therefore remains free outside the calibration phase to take pictures of license plates of vehicles in violation.
- Fig.l general view of the system, which is composed of two cameras (105) and (110) for the stereoscopic recordings and which send video streams to an acquisition card FPGA (185).
- the FPGA card performs real-time video processing and sends the results to the hardened PC (180) via a PCI EXPRESS connector.
- Fig.2 general view of the speed estimation system by video processing, based on stereoscopic calibration.
- Fig.3 rfigure illustrating the general case of road radars filming a vehicle that approaches or moves away from the camera.
- Fig.4 Generic flowchart which represents the different steps of a video processing speed estimation algorithm.
- Fig.5 flowchart which represents the phases of the calibration algorithm of the speed estimation system of our invention.
- FIG. 6 FIG. Which represents the different views of the road following the inclination of the acquisition camera ⁇ 105), as well as the principle of rectification which aims at straightening the road to have a view like the one presented in ( 500).
- Fig.7 top view of the system comprising the two cameras (105) and (110). It represents the distances and the angles ( ⁇ ⁇ ) of horizontal openings (450) and (455) of the two cameras that enter the stereoscopic calculations.
- FIG. 8 profile view of the on-road system with the two cameras (105) and (110), as well as the location of the points (125) on the road and the angles that form with the optical axis of the camera.
- Fig.9 figure which represents the stereoscopic effect on one of the 4 points (125).
- the present invention relates to a passive speed estimation system which presents an automatic calibration method through the use of a secondary camera.
- the acquisition of the images is done by two cameras (105) and (110), the first step (210) consists in making an estimate of the background (Background) of the scene filmed by the camera (105), comes thereafter the step (205) of subtracting each frame with the image of the Background.
- Median filtering (215) is applied to the resulting image of the subtraction to reduce noise and eliminate pixel outliers (hot spots in the image).
- the reinforcement (highlighting) of the zones of the moving objects is done by the morphological operations (220) is more precisely an opening, that is to say an erosion of the related objects of the binary image followed by a dilation, this application aims to strengthen the moving objects (vehicles) and eliminate all the parasites that can be selected by mistake in the detection phase of vehicles (225).
- the vehicle detection phase makes it possible to extract all the active related objects that correspond to the objects searched for (mobile vehicles on the road), the information on these objects will be used to track them on the successive images and thus to estimate their speed. traveling on the road.
- the real coordinates of the movements of the vehicles are generated by a rectification (Fig6-view 1) of the road which makes it possible to have a calibration of the system.
- the calibration of the camera (105) is the objective of our invention.
- the latter is one of the most important aspects in a video processing-based speed estimation system because it determines the accuracy of the measurement.
- the images are represented in 2D, which does not reflect the reality, because we lose the information on the third dimension which is the depth.
- This calibration is made by the stereoscopic calculation from which the use of the second camera (110). This calculation generates a correction matrix that is used in the speed calculation.
- Both cameras are fixed at a height "h" (425) above the road surface, their optical axes are parallel and inclined at an angle ⁇ ) (405) with respect to the road surface.
- the determination of the calibration matrix is done using the stereoscopic effect, the first step is to select four points on the road (125) using an algorithm that chooses easily identifiable points on the two images (eg points on areas with a fortcontrast) (305).
- D Distance separating the object from the camera in profile view (410).
- a The angle (415) between the line that ties the point to the camera and the horizontal axis of the camera.
- the calibration matrix is composed of the coordinates of the detected points as well as their distances relative to the two cameras (105) and (110). Knowing the angle of aperture of the camera ⁇ p v ) (420) and the angle of inclination of the camera with respect to the ground, step (117) the coordinates of each point are determined by a projection of the polar coordinates to the Cartesian coordinates.
- the calibration matrix (118) is used thereafter to have actual displacements in meters and not in pixels, and thereafter to enter the calculation of the speed.
Abstract
The present invention concerns systems (speed detection devices) for passive speed estimation using video processing, which are used, in particular, for detecting traffic violations (speeding).
Description
Système mono-caméra d'estimation de vitesse des véhicules par traitement vidéo avec calibration multi-caméras par effet stéréoscopique Single-camera system for vehicle speed estimation by video processing with multi-camera calibration by stereoscopic effect
DOMAINE DE L'INVENTION FIELD OF THE INVENTION
Cette invention est relative aux systèmes d'estimation de vitesse (cinémomètres) passifs à base de traitement vidéo, qui sont utilisés en particulier pour réaliser des détections d'infraction au code de la route (excès de vitesse). This invention relates to video processing-based passive velocity estimation systems (speedmeters), which are used in particular for detecting violations of the highway code (speeding).
ETAT DE L'ART ANTERIEUR Les systèmes d'estimation de vitesse de véhicules sur route qui existent actuellement appartiennent à l'une des deux catégories suivantes : STATE OF THE PRIOR ART The current vehicle speed estimation systems currently in use belong to one of two categories:
Les systèmes actifs dans lesquels le système émet de l'énergie vers la cible et qui utilise l'énergie réfléchie par celle-ci pour faire l'estimation de la vitesse comme les systèmes à effet doppler (Radar) ou les systèmes à base du Laser (Lidar). Les systèmes passifs qui réalisent l'estimation de vitesse à partir de l'information disponible sans avoir besoin d'emmètre une énergie dans ce but, en particulier les radars à base du traitement vidéo, parmi ceci on distingue les radars utilisant une seule caméra (US 20080166023 Al) et les radars utilisant plusieurs Caméras d'axes parallèles basé sur l'effet Stéréoscopique (EP 1067388 Al). Dans les systèmes video mono-caméra, une caméra numérique filme un véhicule qui se rapproche ou s'éloigne d'elle(Fig.3). Le flux vidéo,subit un certain nombre d'opérations qui permettent de localiser les véhicules en mouvement et d'estimer leurs mouvements sur l'image(Fig.4). Active systems in which the system emits energy to the target and uses the energy reflected by the system to estimate velocity such as Doppler (Radar) systems or Laser-based systems (Lidar). Passive systems that realize the speed estimation from the available information without having to put an energy for this purpose, in particular video processing based radars, among which we distinguish the radars using a single camera ( US 20080166023 Al) and radars using multiple Parallel Axis Cameras based on Stereoscopic effect (EP 1067388 A1). In single-camera video systems, a digital camera films a vehicle that moves closer to or farther from it (Fig.3). The video stream, undergoes a number of operations that locate the moving vehicles and estimate their movements on the image (Fig.4).
Dans un tel flot l'opération de redressement (Fig5) permet d'associer une distance métrique aux pixels du plan de la route. Cette opération nécessite pour être effectuée de disposer des coordonnées métriques d'au moins quatre points sur le plan de déplacement des véhicules afin d'en déduire une matrice de rectification. Celle-ci s'obtient en générale soit grâce à une calibration manuelle (par un opérateur qui détermine la distance réelle sur la route entre
certain points), soit en positionnant sur la chaussée des marquages au sol dont on connaît à priori les coordonnées et qui sont facilement identifiables sur les images. In such a flow, the righting operation (FIG. 5) makes it possible to associate a metric distance with the pixels of the plane of the road. This operation requires to be made to have the metric coordinates of at least four points on the vehicle movement plan to deduce a grinding matrix. This is generally obtained either through a manual calibration (by an operator who determines the actual distance on the road between certain points), or by positioning on the pavement markings on the ground of which we know a priori the coordinates and which are easily identifiable on the images.
Les systèmes multi-cameras sont basés en général sur l'effet stéréoscopique qui permet de calculer la distance entre la cible et l'axe des caméras-Stéréo en mesurant le décalage de l'objet sur l'image gauche et droite prisent en même instant (brevet EP 1067388 Al), ces systèmes permettent de réaliser des estimations de vitesse sans qu'ils soient besoin d'obtenir au préalable des informations métriques sur le plan de la route. Multi-camera systems are usually based on the stereoscopic effect that calculates the distance between the target and the axis of the cameras-Stereo by measuring the shift of the object on the left and right image taken at the same time (EP Patent No. 1067388 A1), these systems make it possible to make speed estimates without the need to obtain prior metric information on the road map.
Notre invention présente l'avantage de permettre de réaliser la calibration du système mono caméra sans faire appel ni à un opérateur extérieur ni à un marquage au sol. Elle peut être réalisée périodiquement (par exemple tous les jours) ce qui permet de s'assurer que le système est calibré en permanence et que des variations de positionnement de la caméra n'ont pas d'impact sur la précision des estimations du système, en outre il présente l'avantage par apport à un système multi-cameras stéréoscopique de n'utiliser que ponctuellement la seconde caméra pour la calibration. La deuxième caméra reste donc libre en dehors de la phase de calibration pour prendre des clicher des plaques d'immatriculation des véhicules en infraction. Our invention has the advantage of allowing the calibration of the single-camera system without calling on an outside operator or a ground marking. It can be carried out periodically (for example every day) which makes it possible to ensure that the system is permanently calibrated and that variations in positioning of the camera have no impact on the accuracy of the estimates of the system. in addition, it has the advantage that a stereoscopic multi-camera system can only occasionally use the second camera for calibration. The second camera therefore remains free outside the calibration phase to take pictures of license plates of vehicles in violation.
BREVE DESCRIPTION DES DIAGRAMMES ET FIGURES BRIEF DESCRIPTION OF DIAGRAMS AND FIGURES
Fig.l : vue générale du système, qui est composé de deux caméras (105) et (110) pour les enregistrements stéréoscopiques et qui envoient des flux vidéo vers une carte d'acquisition FPGA(185). La carte FPGA fait du traitement vidéo en temps réel et envoie les résultats vers le pc durci(180) via un connecteur PCI EXPRESS. Fig.l: general view of the system, which is composed of two cameras (105) and (110) for the stereoscopic recordings and which send video streams to an acquisition card FPGA (185). The FPGA card performs real-time video processing and sends the results to the hardened PC (180) via a PCI EXPRESS connector.
Fig.2 : vue générale du système d'estimation de vitesse par traitement vidéo, basé sur une calibration par effet stéréoscopique. Fig.3 rfigure illustrant le cas général des radars routiers qui filme un véhicule qui se rapproche ou s'éloigne de la caméra. Fig.2: general view of the speed estimation system by video processing, based on stereoscopic calibration. Fig.3 rfigure illustrating the general case of road radars filming a vehicle that approaches or moves away from the camera.
Fig.4 :organigrammegénérique qui représente les différentes étapes d'unalgorithme d'estimation de vitesse à base de traitement vidéo.
Fig.5 rorganigramme qui représente les phases de l'algorithme de calibration du système d'estimation de vitesse de notre invention. Fig.4: Generic flowchart which represents the different steps of a video processing speed estimation algorithm. Fig.5 flowchart which represents the phases of the calibration algorithm of the speed estimation system of our invention.
Fig.6 : figure qui représente les différentes vues de la route suivant l'inclinaison de la caméra d'acquisition{105), ainsi que le principe de rectification qui a pour objectif de redresser les route pour avoir une vue comme celle présentée en (500). FIG. 6: FIG. Which represents the different views of the road following the inclination of the acquisition camera {105), as well as the principle of rectification which aims at straightening the road to have a view like the one presented in ( 500).
Fig.7 : vue de dessus du système comportant les deux caméras (105) et (110). Elle représente les distances et les angles (φν) d'ouvertures horizontaux (450) et (455) des deux caméras qui rentrent dans les calculs stéréoscopiques. Fig.7: top view of the system comprising the two cameras (105) and (110). It represents the distances and the angles (φ ν ) of horizontal openings (450) and (455) of the two cameras that enter the stereoscopic calculations.
Fig.8 : vue de profil du système sur route avec les deux caméras (105) et (110), ainsi que l'emplacement des points (125) sur la route et les angles que forment avec l'axe optique de la caméra. FIG. 8: profile view of the on-road system with the two cameras (105) and (110), as well as the location of the points (125) on the road and the angles that form with the optical axis of the camera.
Fig.9 : figure qui représente l'effet stéréoscopie sur un des 4 point (125). Fig.9: figure which represents the stereoscopic effect on one of the 4 points (125).
EXPOSE DETAILLE DE L'INVENTION La présente invention concerne un système d'estimation de vitesse passif qui présente une méthode de calibration automatique grâce à l'utilisation d'une caméra secondaire. DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a passive speed estimation system which presents an automatic calibration method through the use of a secondary camera.
Nous décrivons ci-dessous tout d'abord les étapes classiques d'un cinémomètre vidéo (Fig. 4) puis nous détaillons les étapes de la calibration du système qui permettent d'obtenir les informations métriques de 4 points sur la route. Etapes classiques d'un cinémomètre vidéo. We first describe the typical steps of a video speedometer (Fig. 4) and then detail the system calibration steps that provide the metric information of 4 points on the road. Classic steps of a video speedometer.
L'acquisition des images se fait par deux caméras (105) et (110), la première étape (210)consiste à faire une estimation de l'arrière-plan (Background) de la scène filmée par la caméra (105), vient par la suite l'étape (205) de soustraction de chaque frame avec l'image du Background. Le filtrage médian (215) est appliqué sur l'image résultante de la soustraction afin de réduire le bruit et d'éliminer les valeurs aberrantes des pixels (points chauds dans l'image).
Le renforcement (mise en évidence) des zones des objets mobiles se fait par les opérations morphologiques (220) est plus précisément une ouverture, c'est-à-dire une érosion des objets connexes de l'image binaire suivie d'une dilatation, cette application a pour objectif de renforcer les objets mobiles (véhicules) et d'éliminer tous les parasites qui peuvent être sélectionnés par erreur à la phase de détection des véhicules (225). The acquisition of the images is done by two cameras (105) and (110), the first step (210) consists in making an estimate of the background (Background) of the scene filmed by the camera (105), comes thereafter the step (205) of subtracting each frame with the image of the Background. Median filtering (215) is applied to the resulting image of the subtraction to reduce noise and eliminate pixel outliers (hot spots in the image). The reinforcement (highlighting) of the zones of the moving objects is done by the morphological operations (220) is more precisely an opening, that is to say an erosion of the related objects of the binary image followed by a dilation, this application aims to strengthen the moving objects (vehicles) and eliminate all the parasites that can be selected by mistake in the detection phase of vehicles (225).
La phase de détection des véhicules permet d'extraire tous les objets connexes actifs qui correspondent aux objets cherchés (des véhicules mobiles sur route), les informations sur ces objets vont être utilisées pour en faire le suivi sur les images successives et ainsi estimer leur vitesse de déplacement sur la route. Les coordonnés réels des déplacements des véhicules sont générés par une rectification (Fig6-vue 1) de la route qui permet d'avoir une calibration du système. The vehicle detection phase makes it possible to extract all the active related objects that correspond to the objects searched for (mobile vehicles on the road), the information on these objects will be used to track them on the successive images and thus to estimate their speed. traveling on the road. The real coordinates of the movements of the vehicles are generated by a rectification (Fig6-view 1) of the road which makes it possible to have a calibration of the system.
Une fois le calcul des déplacements réel fait, l'étape finale (235)consiste à calculer la vitesse des véhicules par la formule suivante : v ( m ) = (d2 - dl). (fps). (3,6) (d2-dl) :est le déplacement métrique du véhicule. fps : est le nombre d'image par seconde. Technique de calibration par stéréoscopie Once the actual displacement calculation is done, the final step (235) is to calculate the speed of the vehicles by the following formula: v ( m ) = (d2 - dl). (Fps). (3,6) (d2-dl): is the metric displacement of the vehicle. fps: is the frame number. Stereoscopic calibration technique
La calibration de la caméra (105) est l'objectif de notre invention. Cette dernière est l'un des aspects les plus importants dans un système d'estimation de vitesse à base de traitement vidéo car elle conditionne la précision de la mesure. Les images étant représentées en 2D, ce qui ne reflète pas la réalité, car nous perdons l'information sur la troisième dimension qui est la profondeur. D'où l'intérêt de faire une rectification de l'image ou une calibration métrique afin de faire correspondre chaque pixel à une coordonnée métrique représentant l'emplacement de ce pixel sur l'image via ses coordonnées x et y sur la route. Cette calibration est faite par le calcul stéréoscopique d'où l'utilisation de la deuxième caméra (110). Ce calcul génère une matrice de rectification qui est utilisée dans le calcul de vitesse.
Les deux caméras sont fixées à une hauteur "h" (425) au-dessus de la surface de la route, leurs axes optiques sontparallèles et inclinés à un angle { Θ ) (405) par rapport à la surface de la route. The calibration of the camera (105) is the objective of our invention. The latter is one of the most important aspects in a video processing-based speed estimation system because it determines the accuracy of the measurement. The images are represented in 2D, which does not reflect the reality, because we lose the information on the third dimension which is the depth. Hence the interest of making an image correction or a metric calibration to match each pixel to a metric coordinate representing the location of this pixel on the image via its x and y coordinates on the road. This calibration is made by the stereoscopic calculation from which the use of the second camera (110). This calculation generates a correction matrix that is used in the speed calculation. Both cameras are fixed at a height "h" (425) above the road surface, their optical axes are parallel and inclined at an angle {Θ) (405) with respect to the road surface.
La détermination de la matrice de calibration se fait en utilisant l'effet stéréoscopique, la première étape est de sélectionnerquatre pointssur la route (125) en utilisant un algorithme qui choisitdes pointsfacilement identifiable sur les deux images (par exemple des points sur des zones avec un fortcontraste) (305). The determination of the calibration matrix is done using the stereoscopic effect, the first step is to select four points on the road (125) using an algorithm that chooses easily identifiable points on the two images (eg points on areas with a fortcontrast) (305).
Une fois qu'on localise les quatre points(125),on passe au calcul des distances (410) séparant ces points avec l'axe (115) des deux caméras (105) et (110) en utilisant la formule de distance stéréoscopique (voir les paramètres sur les figures Fig7et Fig8).
Once we locate the four points (125), we proceed to the calculation of the distances (410) separating these points with the axis (115) of the two cameras (105) and (110) using the stereoscopic distance formula ( see the parameters in the figures Fig7and Fig8).
2 tan(Ç) . (xc_xD ) 2 tan ()). (x c _x D )
B : La distance entre les deux caméras (120). <ph : L'ouverture de la caméra (450). B: The distance between the two cameras (120). <p h : Opening the camera (450).
(XG-XD ) : La distance (560) entre les centres (510) et (520)en pixels de point (125) sur les deux images. (XG-XD) : The distance (560) between centers (510) and (520) in dot pixels (125) on both images.
D D
d = — d = -
cos(a) d : Distance (400) latérale entre la caméra et le point (125). cos (a) d: The lateral distance (400) between the camera and the point (125).
D : Distance séparant l'objet de la caméra en vue de profil (410). a : L'angle(415) qui sépare la droite qui fait lier le point à la caméra et l'axe horizontale de la caméra. La matrice de calibration est composée des coordonnées des points détectés ainsi que leurs distances par rapports aux deux caméras (105) et (110). En connaissant l'angle d'ouverture de la caméra { <pv ) (420) et l'angle d'inclinaison de la caméra par rapport au sol, l'étape (117)
de détermination des coordonnées de chaque point se fait par une projection des coordonnées polaires aux coordonnées cartésiennes. D: Distance separating the object from the camera in profile view (410). a: The angle (415) between the line that ties the point to the camera and the horizontal axis of the camera. The calibration matrix is composed of the coordinates of the detected points as well as their distances relative to the two cameras (105) and (110). Knowing the angle of aperture of the camera {<p v ) (420) and the angle of inclination of the camera with respect to the ground, step (117) the coordinates of each point are determined by a projection of the polar coordinates to the Cartesian coordinates.
La matrice de calibration (118) est utilisé par la suite pour avoir des déplacements réels en mètre et non pas en pixels, et par la suite la faire rentrer au calcul de la vitesse.
The calibration matrix (118) is used thereafter to have actual displacements in meters and not in pixels, and thereafter to enter the calculation of the speed.
Claims
1. Système d'estimation des vitesses d'objets par traitement vidéo caractérisé en ce que le système est composé d'une caméra principale (105) de l'estimation de vitesse et d'une deuxième caméra de calibration (110). Les deux caméras sont fixées à une hauteur "h" (400) au-dessus de la surface de la route, leurs axes optiques sont parallèles avec une distance (B) les séparant et elles sont inclinés à un angle # (405) par rapport à la surface de la route. Sur les images prises par les deux caméras au même instant t, une unité de traitement (190) permettant l'indentification des points de références (125) pour la calibration ainsi que leurs distances réelles métriques par rapport au sol. Une matrice de calibration composée des coordonnées des points identifiés et de leurs distances métriques d (400) est utilisée par une unité de traitement (190) pour estimer la vitesse de mouvement. 1. System for estimating the speeds of objects by video processing, characterized in that the system is composed of a main camera (105) of the speed estimation and a second calibration camera (110). The two cameras are fixed at a height "h" (400) above the road surface, their optical axes are parallel with a distance (B) separating them and they are inclined at an angle # (405) relative to each other. on the surface of the road. On the images taken by the two cameras at the same time t, a processing unit (190) for the identification of the reference points (125) for the calibration and their actual metric distances from the ground. A calibration matrix composed of the coordinates of the identified points and their metric distances d (400) is used by a processing unit (190) to estimate the speed of movement.
2. Système selon la revendication 1 caractérisé en ce que les caméras (105) et (110) sont numériques et identiques avec des axes optiques parallèles. 2. System according to claim 1 characterized in that the cameras (105) and (110) are digital and identical with parallel optical axes.
3. Système multi-caméras selon les revendications 1 et 2 caractérisé en ce que l'unité de traitement des images (200) contient une carte d'acquisition (210) qui permet de récupérer le flux vidéo en provenance des caméras (105) et (110), identifier les éléments mobiles, et calibrer la scène filmée. 3. Multi-camera system according to claims 1 and 2 characterized in that the image processing unit (200) contains an acquisition card (210) for recovering the video stream from the cameras (105) and (110), identify the moving elements, and calibrate the filmed scene.
4. Système multi-caméras selon les revendications 1 à 3 caractérisé en ce qu'au niveau de l'unité de traitement, un algorithme permet d'identifier 4 points (de fortes contraste par exemple) sur les deux images acquises par les deux caméras (105) et (110) afin de calibrer le système et de réaliser le redressement de la scène en utilisant l'effet stéréoscopique. 4. Multi-camera system according to claims 1 to 3 characterized in that at the processing unit, an algorithm can identify 4 points (high contrast for example) on the two images acquired by the two cameras (105) and (110) to calibrate the system and perform scene recovery using the stereoscopic effect.
5. Système multi-caméras selon les revendications 1 à 4 caractérisé en ce que la caméra (105) est une caméra principale qui a pour objectif la détection des objets mobiles sur la scène filmée, et que la caméra (110) est une caméra de calibration qui permet de redressé la scène par l'effet stéréoscopique. 5. Multi-camera system according to claims 1 to 4 characterized in that the camera (105) is a main camera that aims to detect moving objects on the scene filmed, and that the camera (110) is a camera of calibration that rectifies the scene by the stereoscopic effect.
6. Système multi-caméras selon les revendications 1 à 5 caractérisé en ce que les deux caméras peuvent réaliser un système de calibrage croisé, c'est-à-dire que chaque caméra peut prendre le rôle principal de détection des objets mobile et calcul de vitesse comme elle peut prendre le rôle de calibrage de la deuxième caméra.
6. Multi-camera system according to claims 1 to 5 characterized in that the two cameras can realize a cross calibration system, that is to say that each camera can take the main role of mobile objects detection and calculation of speed as it can take the role of calibrating the second camera.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
MA36097A MA36097B1 (en) | 2013-07-08 | 2013-07-08 | Mono-camera system for estimating vehicle speed by video processing with multi-camera calibration by stereoscopic effect |
MA36097 | 2013-07-08 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2015005758A1 true WO2015005758A1 (en) | 2015-01-15 |
WO2015005758A4 WO2015005758A4 (en) | 2015-03-19 |
Family
ID=51429344
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/MA2014/000016 WO2015005758A1 (en) | 2013-07-08 | 2014-07-08 | Single-camera system for estimating the speed of vehicles by video processing with multi-camera calibration by stereoscopic effect |
Country Status (2)
Country | Link |
---|---|
MA (1) | MA36097B1 (en) |
WO (1) | WO2015005758A1 (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5809161A (en) * | 1992-03-20 | 1998-09-15 | Commonwealth Scientific And Industrial Research Organisation | Vehicle monitoring system |
US5910817A (en) * | 1995-05-18 | 1999-06-08 | Omron Corporation | Object observing method and device |
US20090207046A1 (en) * | 2006-03-22 | 2009-08-20 | Kria S.R.L. | system for detecting vehicles |
-
2013
- 2013-07-08 MA MA36097A patent/MA36097B1/en unknown
-
2014
- 2014-07-08 WO PCT/MA2014/000016 patent/WO2015005758A1/en active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5809161A (en) * | 1992-03-20 | 1998-09-15 | Commonwealth Scientific And Industrial Research Organisation | Vehicle monitoring system |
US5910817A (en) * | 1995-05-18 | 1999-06-08 | Omron Corporation | Object observing method and device |
US20090207046A1 (en) * | 2006-03-22 | 2009-08-20 | Kria S.R.L. | system for detecting vehicles |
Also Published As
Publication number | Publication date |
---|---|
MA20150045A1 (en) | 2015-02-27 |
MA36097B1 (en) | 2016-01-29 |
WO2015005758A4 (en) | 2015-03-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6899369B2 (en) | Methods and systems for generating and using localization reference data | |
US8238610B2 (en) | Homography-based passive vehicle speed measuring | |
JP6726932B2 (en) | Context-aware object detection in aerial photography/video using road metadata | |
US20170337434A1 (en) | Warning Method of Obstacles and Device of Obstacles | |
US8571354B2 (en) | Method of and arrangement for blurring an image | |
US10909395B2 (en) | Object detection apparatus | |
EP2963634B1 (en) | Stereo camera device | |
CN111754581A (en) | Camera calibration method, roadside sensing equipment and intelligent traffic system | |
KR20160062880A (en) | road traffic information management system for g using camera and radar | |
WO2016185373A1 (en) | Detection and documentation of tailgating and speeding violations | |
KR101751112B1 (en) | Apparatus for generating traffic information base on image and Method thereof | |
KR101431373B1 (en) | Apparatus for estimating of vehicle movement using stereo matching | |
CN111213153A (en) | Target object motion state detection method, device and storage medium | |
US10832428B2 (en) | Method and apparatus for estimating a range of a moving object | |
KR101709317B1 (en) | Method for calculating an object's coordinates in an image using single camera and gps | |
US20180182113A1 (en) | Image processing device and image processing method | |
CN111353453A (en) | Obstacle detection method and apparatus for vehicle | |
CN108399360B (en) | Continuous obstacle detection method, device and terminal | |
KR102195040B1 (en) | Method for collecting road signs information using MMS and mono camera | |
WO2015005758A1 (en) | Single-camera system for estimating the speed of vehicles by video processing with multi-camera calibration by stereoscopic effect | |
CN109345576A (en) | Vehicle Speed identification method and system | |
CN115249345A (en) | Traffic jam detection method based on oblique photography three-dimensional live-action map | |
JP6786635B2 (en) | Vehicle recognition device and vehicle recognition method | |
Thai et al. | Application of edge detection algorithm for self-driving vehicles | |
Kaneko et al. | Vehicle speed estimation by in-vehicle camera |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14757988 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14757988 Country of ref document: EP Kind code of ref document: A1 |