Abstract: A method for detecting an incident on a traffic lane uses a video camera having a target constituting an optoelectronic converter of a real optical image, the target being monitored by a programmable processing member and the incident detection process being designed to be carried out by activating the programmable processing member only when the real landscape image focused on the target is stationary. The inventive method includes detecting the beginning of displacement of the real image of the landscape relative to the target, in deactivating the programmable processing member right at the beginning of the displacement of the real image of the landscape relative to the target, and reactivating the programmable processing member at the end of the displacement of the real image of the landscape relative to the target to carry out the incident detection process.
Abstract: Method and device for stabilizing images obtained by a video camera of an environment with objects moving along lanes. The method a first stage of determining in first images obtained by the camera, portions corresponding to the movement lanes, determining the remaining second portions complementary to the first portions, determining, in the second portions, the reference position of the image point corresponding to a first stationary object point, and a second stage of determining in at least one second image taken after the first images, the vector that corresponds to the movement of the image point relative to its reference position determined by the first images, and processing the second image as a function of the modulus and direction of the vector such that, relative to the three-dimensional frame of reference the image point returns to its reference position.
Abstract: Methods for calibrating at least two video cameras for a stereoscopic device includes: providing, on the lane portion, nine marks of hue other than that of the lane, sequenced on a first set of three concurrent virtual straight lines in a first point and distributed in specific manner on a second set of concurrent virtual straight lines in a second point; forming, with each of the cameras an image of the lane portion; defining, in each of the two images, one characteristic point of each mark image; determining, with the characteristic points, six concurrent straight line images respectively in two concurrent points; and processing the video signals delivered by each video camera such that the signals are representative of two images suitable for forming a stereoscopic video image. The method is useful for determining the occupancy condition of a lane portion and detecting incidents.
Type:
Grant
Filed:
July 9, 2004
Date of Patent:
July 7, 2009
Assignee:
Citilog
Inventors:
Jerome Douret, Ryad Benosman, Jean Devars, Salah Bouzar
Abstract: Method and device for stabilizing images obtained by a video camera of an environment with objects moving along lanes. The method a first stage of determining in first images obtained by the camera, portions corresponding to the movement lanes, determining the remaining second portions complementary to the first portions, determining, in the second portions, the reference position of the image point corresponding to a first stationary object point, and a second stage of determining in at least one second image taken after the first images, the vector that corresponds to the movement of the image point relative to its reference position determined by the first images, and processing the second image as a function of the modulus and direction of the vector such that, relative to the three-dimensional frame of reference the image point returns to its reference position.
Abstract: The invention concerns a method for detecting an incident on a traffic lane portion (1) located in a landscape (2) when the lane portion is likely to be used by objects. The invention is characterized in that the method uses a video camera (3) comprising a target (4) constituting an optoelectronic converter of a real optical image of the landscape (2), the target being monitored by a programmable processing member (6) and the incident detection process being designed to be carried out by activating the programmable processing member (6) only when the real landscape image focused on the target (4) is stationary.