Range discontinuities (44) in an imaged scene are located with superior accuracy by using a pair of stereoscopically related images (10, 12) of the scene to sense edges in the scene, and to determine the range of correlatable image pixels; segmenting the scene into regions (36) of similar intensity; assigning to each region (36) a range based upon the range of any correlatable pixels in it; and locating the boundary (44) between adjacent regions of substantially dissimilar range.
A method is disclosed for generating a categorical depth map of a scene using passive defocus sensing. In a preferred embodiment three synchronized CCD cameras focused at different distances detect three images of the same scene. An image processor partitions the images into an array of regions and calculates a sharpness value for each region. The sharpness value for a region is calculated by summing over all pixels (x,y) in the region the absolute difference in the intensity value of a pixel (x,y)( with pixel (x-k,y-l), where k and l are constants. The image processor then constructs a depth map of the scene by determining for each region the image with the greatest sharpness in that region. An application of the invention to a mobile robot control system is described in detail. Among other applications, the method may be used for collision avoidance, object detection, and speed measurement.
An image processor (10) is provided in which imaging data obtained by an image sensor (12) is processed for purposes of detecting motion. An optical flow field generator (16) processes the image data to provide condition number-based smoothed optical flow vectors, and stores those smoothed vectors in memory (18). The optical flow field stored in memory (18) is accessed and further processed by analyzer (20) for purposes of providing users with information on the motion of objects in the image plane of image sensor (12). Such user information may be displayed on a monitor (22).
A method and system are provided for relating a passive sensor (14) to a geographic environment (42, 44). The passive sensor (14) senses an image (40) of the geographic environment (42, 44). At least one feature (46) is extracted (26) from the image (40). At least one feature is generated (30) from map data (28) representative of the geographic environment. At least one extracted feature is related (22, 32) to at least one generated feature, such that the passive sensor (14) is related to the geographic environment (42, 44).
A method and system are provided for relating a passive sensor (14) to a geographic environment (42, 44). The passive sensor (14) senses an image (40) of the geographic environment (42, 44). At least one feature (46) is extracted (26) from the image (40). At least one feature is generated (30) from map data (28) representative of the geographic environment. At least one extracted feature is related (22, 32) to at least one generated feature, such that the passive sensor (14) is related to the geographic environment (42, 44).
A three-dimensional measuring method obtains three-dimensional information related to an image including one or a plurality of objects based on input information which describes the image. The method includes the steps of (a) selecting information necessary for calculating marks of presumed information from the input information, where each of the marks are used to evaluate the appropriateness of the presumed information, (b) calculating marks of each pixel of the image based on the information selected by the step (a) and based on feedback information which is derived from an internal state, where the internal state includes information related to the object for each pixel of the image, (c) adding the marks for each pixel calculated in the step (b) for each object so as to obtain a total mark, (d) arbitrarily changing the internal state so as to reduce the total mark, and (e) outputting the internal state in which the total mark is a local minimum as the three-dimensional information.
