Abstract: The present invention provides a computer implemented process for detecting multi-view multi-pose objects. The process comprises training of a classifier for each intra-class exemplar, training of a strong classifier and combining the individual exemplar-based classifiers with a single objective function. This function is optimized using the two nested AdaBoost loops. The first loop is the outer loop that selects discriminative candidate exemplars. The second loop, the inner loop selects the discriminative candidate features on the selected exemplars to compute all weak classifiers for a specific position such as a view/pose. Then all the computed weak classifiers are automatically combined into a final classifier (strong classifier) which is the object to be detected.

Abstract: A system and method for turning a regular computer monitor screen into a touch screen using an ordinary camera. It includes an image-screen mapping procedure to correct for the non-flatness of the computer screen. It also includes a segmentation method to distinguish the foreground, for example an indicator such as a finger, from the background of a computer screen. Additionally, this system and method includes a robust technique of finding the tip point location of the indicator (such as the finger tip). The screen coordinates of the tip point are then used to control the position of the system indicator.

Abstract: The present invention provides an improved system and method for object detection with histogram of oriented gradient (HOG) based support vector machine (SVM). Specifically, the system provides a computational framework to stably detect still or not moving objects over a wide range of viewpoints. The framework includes providing a sensor input of images which are received by the “focus of attention” mechanism to identify the regions in the image that potentially contain the target objects. These regions are further computed to generate hypothesized objects, specifically generating selected regions containing the target object hypothesis with respect to their positions. Thereafter, these selected regions are verified by an extended HOG-based SVM classifier to generate the detected objects.

Abstract: An incremental motion estimation system and process for estimating the camera pose parameters associated with each image of a long image sequence. Unlike previous approaches, which rely on point matches across three or more views, the present system and process also includes those points shared only by two views. The problem is formulated as a series of localized bundle adjustments in such a way that the estimated camera motions in the whole sequence are consistent with each other. The result of the inclusion of two-view matching points and the localized bundle adjustment approach is more accurate estimates of the camera pose parameters for each image in the sequence than previous incremental techniques, and providing an accuracy approaching that of global bundle adjustment techniques except with processing times about 100 to 700 times faster than the global approaches.

Abstract: Method and a system for correlating pixels between two digital images. In general, the present invention uses a progressive iterative technique that finds generally unambiguous pixel matches by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches in the current iteration then are found using the correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique of the present invention that provides a more robust estimate of pixel match reliability. Potential pixel matches found in the search ranges are tested for ambiguity and any unambiguous matches are selected and added to the set of reliable pixel matches. The ambiguity testing includes determining a correlation score for the pixel match and classifying the match based on the correlation score.

Abstract: A system and method, called Image-Based Surface Detail Transfer, to transfer geometric details from one surface of an object in an image to another with simple 2D image operations. The basic observation is that, without knowing its 3D geometry, geometric details (local deformations) can be extracted from a single image of an object in a way independent of its surface reflectance, and furthermore, these geometric details can be transferred to modify the appearance of other objects directly in images. Examples are shown including surface detail transfer between real objects, as well as between real and synthesized objects.

Abstract: A method and a system for progressive stereo matching of digital images representing a scene. In general, the present invention uses a progressive iterative technique that includes a disparity gradient limit principle and a least commitment strategy. Generally unambiguous pixel matches are found by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches are used to define the search ranges for each pixel to guide matching in the current iteration. Unambiguous pixel matches then are found using a novel correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique that provides a more robust estimate of pixel match reliability. Potential pixel matches found in the search ranges are tested for ambiguity and unambiguous matches are selected and added to the reliable pixel matches.

Abstract: A system and method for identifying objects, particularly vehicles, between two non-overlapping cameras. More specifically, the method and system determines whether a vehicle depicted in an image captured by a first camera is the same vehicle or a different vehicle than a vehicle depicted in an image captured by a second camera. This intra-camera analysis determines whether the vehicle viewed by the first camera is the same as the vehicle viewed by the second camera, without directly matching the two vehicle images, thus eliminating the problems and inaccuracies caused by disparate environmental conditions acting on the two cameras, such as dramatic appearance and aspect changes.

Abstract: A method and apparatus for tracking a movable object using a plurality of images, each of which is separated by an interval of time is disclosed. The plurality of images includes first and second images. The method and apparatus include elements for aligning the first and second images as a function of (i) at least one feature of a first movable object captured in the first image, and (ii) at least one feature of a second movable object captured in the second image; and after aligning the first and second images, comparing at least one portion of the first image with at least one portion of the second image.

Abstract: A method and a system for progressive stereo matching of digital images representing a scene. In general, the present invention uses a progressive iterative technique that includes a disparity gradient limit principle and a least commitment strategy. Generally unambiguous pixel matches are found by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches are used to define the search ranges for each pixel to guide matching in the current iteration. Unambiguous pixel matches then are found using a novel correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique that provides a more robust estimate of pixel match reliability. Potential pixel matches found in the search ranges are tested for ambiguity and unambiguous matches are selected and added to the reliable pixel matches.

Abstract: A method and apparatus for unsupervised learning of measures for matching objects between images from at least two non-overlapping cameras is disclosed The method includes collecting at least one two pairs of feature maps, where the at least one two pairs of feature maps are derived from features of objects captured in the images. The method further includes computing, as a function of at least one two pairs of feature maps, at least one first and second match measures, wherein the first match measure is of a same class and the second match measure is of a different class.

Abstract: A hinge device with a locking function includes a first locking element, a first connecting element, a second locking element and a connecting shaft. The first locking element has a first through hole and a first locking portion. The first connecting element has a second through hole. The first connecting element is connected with the first locking element. The second locking element is disposed between the first through hole and the second through hole. The second locking element has a shaft hole and a second locking portion corresponding to the first locking portion. The connecting shaft passes through the first through hole, the shaft hole and the second through hole so as to form the hinge device.

Abstract: A system and method, called Image-Based Surface Detail Transfer, to transfer geometric details from one surface of an object in an image to another with simple 2D image operations. The basic observation is that, without knowing its 3D geometry, geometric details (local deformations) can be extracted from a single image of an object in a way independent of its surface reflectance, and furthermore, these geometric details can be transferred to modify the appearance of other objects directly in images. Examples are shown including surface detail transfer between real objects, as well as between real and synthesized objects.

Abstract: An incremental motion estimation system and process for estimating the camera pose parameters associated with each image of a long image sequence. Unlike previous approaches, which rely on point matches across three or more views, the present system and process also includes those points shared only by two views. The problem is formulated as a series of localized bundle adjustments in such a way that the estimated camera motions in the whole sequence are consistent with each other. The result of the inclusion of two-view matching points and the localized bundle adjustment approach is more accurate estimates of the camera pose parameters for each image in the sequence than previous incremental techniques, and providing an accuracy approaching that of global bundle adjustment techniques except with processing times about 100 to 700 times faster than the global approaches.

Abstract: A method and apparatus for recognizing an object, comprising providing a set of scene features from a scene, pruning a set of model features, generating a set of hypotheses associated with the pruned set of model features for the set of scene features, pruning the set of hypotheses, and verifying the set of pruned hypotheses is provided.

Abstract: An incremental motion estimation system and process for estimating the camera pose parameters associated with each image of a long image sequence. Unlike previous approaches, which rely on point matches across three or more views, the present system and process also includes those points shared only by two views. The problem is formulated as a series of localized bundle adjustments in such a way that the estimated camera motions in the whole sequence are consistent with each other. The result of the inclusion of two-view matching points and the localized bundle adjustment approach is more accurate estimates of the camera pose parameters for each image in the sequence than previous incremental techniques, and providing an accuracy approaching that of global bundle adjustment techniques except with processing times about 100 to 700 times faster than the global approaches.

Abstract: A system and method, called Image-Based Surface Detail Transfer, to transfer geometric details from one surface of an object in an image to another with simple 2D image operations. The basic observation is that, without knowing its 3D geometry, geometric details (local deformations) can be extracted from a single image of an object in a way independent of its surface reflectance, and furthermore, these geometric details can be transferred to modify the appearance of other objects directly in images. Examples are shown including surface detail transfer between real objects, as well as between real and synthesized objects.

Abstract: A method and a system for progressive stereo matching of digital images representing a scene. In general, the present invention uses a progressive iterative technique that includes a disparity gradient limit principle and a least commitment strategy. The progressive iterative technique finds generally unambiguous pixel matches by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches found in the previous iteration are used to define the search ranges for each pixel to guide matching in the current iteration. Unambiguous pixel matches in the current iteration then are found using a novel correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique of the present invention that provides a more robust estimate of pixel match reliability.

Abstract: A method and a system for progressive stereo matching of digital images representing a scene. In general, the present invention uses a progressive iterative technique that includes a disparity gradient limit principle and a least commitment strategy. Generally unambiguous pixel matches are found by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches are used to define the search ranges for each pixel to guide matching in the current iteration. Unambiguous pixel matches then are found using a novel correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique that provides a more robust estimate of pixel match reliability. Potential pixel matches found in the search ranges are tested for ambiguity and unambiguous matches are selected and added to the reliable pixel matches.

Abstract: A method and a system for progressive stereo matching of digital images representing a scene. In general, the present invention uses a progressive iterative technique that includes a disparity gradient limit principle and a least commitment strategy. Generally unambiguous pixel matches are found by beginning with a few reliable pixel matches and finding progressively more unambiguous pixel matches. Unambiguous pixel matches are used to define the search ranges for each pixel to guide matching in the current iteration. Unambiguous pixel matches then are found using a novel correlation technique and based on a correlation score associated with a pixel match. The search range is capable of being rotated, and is part of a novel correlation technique that provides a more robust estimate of pixel match reliability. Potential pixel matches found in the search ranges are tested for ambiguity and unambiguous matches are selected and added to the reliable pixel matches.