Abstract: A method and an apparatus for detecting an abnormal situation are disclosed. The method includes detecting whether a first target exists in an obtained image; recognizing whether the first target holds an object, when the first target exists in the image; obtaining motion information of the object, when the first target holds the object; and determining, based on the motion information of the object, whether the abnormal situation exists.

Abstract: A method and an apparatus for detecting an abnormal situation are disclosed. The method includes recognizing whether a detection target exists in a captured image; generating, based on the captured image, a three-dimensional point cloud of the detection target in the captured image, when the detection target exists; obtaining, based on the generated three-dimensional point cloud, one or more current posture features of the detection target; and determining, based on the current posture features and one or more predetermined posture feature standards, whether the abnormal situation exists, the posture feature standards being previously determined based on one or more common features when the detection target performs a plurality of abnormal actions.

Abstract: A method and an apparatus for separating objects are disclosed. The method includes obtaining a depth image including a plurality of objects; obtaining a two-dimensional image including the objects; performing pixel-clustering using depth values of pixels in the depth image and pixel values of pixels in the two-dimensional image to obtain a plurality of sub-regions; performing region-clustering for the sub-regions to obtain a clustering result as an object separation result; and outputting the object separation result.

Abstract: A method and an apparatus for detecting persons are disclosed. The method includes initially detecting the persons in a height-top-view; dividing the height-top-view into one or more regions, and estimating crowd density in each region; determining, based on the crowd density, visible regions of the initially detected persons in each of the regions; for each of the initially detected persons, extracting a first gradient feature and a second gradient feature of the person from the height-top-view, and a grayscale image or a color image corresponding to the height-top-view, respectively; for each of the initially detected persons, determining, based on the extracted first gradient feature and second gradient feature, using a previously constructed classifier corresponding to the determined visible region of the person, a confidence level of the initially detected person; and correcting, based on the confidence level, a detection result of the initially detected persons.

Abstract: A method and an apparatus for detecting an interest degree of a crowd in a target position are disclosed. The interest degree detection method includes projecting a depth image obtained by photographing onto a height-top-view, the depth image including the crowd and the target position; dividing the height-top-view into cells; determining density of the crowd in each cell; determining a moving speed and a moving direction of the crowd in each cell; determining orientation of the crowd in each cell; and determining, based on the density, the moving speed, the moving direction and the orientation of the crowd, the interest degree of the crowd in each cell in the target position. According to this method, the interest degree of the crowd in the target position can be detected accurately, even at a crowded place where it is difficult to detect and track a single person.

Abstract: Disclosed is an object dangerousness recognition method comprising steps of generating, based on an image captured by a stereo camera, a heterogeneous point cloud of an object in the image, each point in the heterogeneous point cloud having depth information and planar image information; determining, based on the depth information and the planar image information of each point in the heterogeneous point cloud, a solid shape of the object, and then, generating a first dangerousness parameter according to the solid shape; determining, based on the depth information and the planar image information of each point in the heterogeneous point cloud, a surface feature of the object, and then, generating a second dangerousness parameter according to the surface feature; and generating, based on the first and second dangerousness parameters, a comprehensive dangerousness parameter of the object.

Abstract: Method, device, and non-transitory computer-readable medium detecting a gathering of objects based on stereo vision are disclosed, and the method comprises steps of obtaining current and prior images and a corresponding depth map; extracting foreground pixels corresponding to detection objects from the current and prior images, and projecting the foreground pixels onto a ground surface to acquire a foreground projection image including foreground projection blocks; conducting, based on image feature differences of the foreground pixels between the current and prior images, projection onto the ground surface to acquire moving foreground projection blocks; utilizing the moving foreground projection blocks to erode the foreground projection blocks to obtain still foreground projection blocks; and determining, based on the still foreground projection blocks, whether the gathering of objects exists.

Abstract: A method and an apparatus for detecting an object are disclosed. The method comprises the steps of obtaining a plurality of depth images of the object; extracting foregrounds; fusing the foregrounds in a unified three-dimensional world coordinate system; calculating an appearance two-dimensional histogram from the fused foreground by the following steps of dividing the foreground into vertical members and getting statistics of the numbers of foreground points in the vertical members so as to obtain the appearance two-dimensional histogram; determining an overlapping region and the number of overlaps based on the placement of stereo cameras; determining a detection parameter relating to a detection position based on the overlapping region and the number of overlaps; and detecting the object by the appearance two-dimensional histogram based on the determined detection parameter.

Abstract: A method and a system for updating a background model based on depth are disclosed. The method includes receiving, in response to the occurrence of a predetermined background updating condition, one or more depth images captured after a time when the predetermined background updating condition occurs; obtaining, based on an original background model, foreground images in the one or more captured depth images, which are newly added compared with a depth image at the time when the predetermined background updating condition occurs; for each of foreground pixels in each of the newly added foreground images, comparing a current depth value with a previous depth value before the time when the predetermined background updating condition occurs; and updating, when the current depth value is greater than the previous depth value, the original background model as the updated background model by using the foreground pixel in the newly added foreground image.

Abstract: Disclosed are an object positioning method and an object positioning device based on object detection results of plural stereo cameras. The method comprises a step of obtaining, when each of the plural stereo cameras continuously carries out tracking and detection with respect to each of objects, positional information of the corresponding object; a step of generating, based on the positional information of the corresponding object, a trajectory of the corresponding object; and a step of carrying out a merging process with respect to the trajectories generated corresponding to the plural stereo cameras so as to determine at least one object position.

Abstract: A method and an apparatus for separating objects are disclosed. The method includes obtaining a depth image including a plurality of objects; obtaining a two-dimensional image including the objects; performing pixel-clustering using depth values of pixels in the depth image and pixel values of pixels in the two-dimensional image to obtain a plurality of sub-regions; performing region-clustering for the sub-regions to obtain a clustering result as an object separation result; and outputting the object separation result.

Abstract: A method and an apparatus for detecting an abnormal situation are disclosed. The method includes recognizing whether a detection target exists in a captured image; generating, based on the captured image, a three-dimensional point cloud of the detection target in the captured image, when the detection target exists; obtaining, based on the generated three-dimensional point cloud, one or more current posture features of the detection target; and determining, based on the current posture features and one or more predetermined posture feature standards, whether the abnormal situation exists, the posture feature standards being previously determined based on one or more common features when the detection target performs a plurality of abnormal actions.

Abstract: Method, device, and non-transitory computer-readable medium detecting a gathering of objects based on stereo vision are disclosed, and the method comprises steps of obtaining current and prior images and a corresponding depth map; extracting foreground pixels corresponding to detection objects from the current and prior images, and projecting the foreground pixels onto a ground surface to acquire a foreground projection image including foreground projection blocks; conducting, based on image feature differences of the foreground pixels between the current and prior images, projection onto the ground surface to acquire moving foreground projection blocks; utilizing the moving foreground projection blocks to erode the foreground projection blocks to obtain still foreground projection blocks; and determining, based on the still foreground projection blocks, whether the gathering of objects exists.

Abstract: A method and an apparatus for detecting persons are disclosed. The method includes initially detecting the persons in a height-top-view; dividing the height-top-view into one or more regions, and estimating crowd density in each region; determining, based on the crowd density, visible regions of the initially detected persons in each of the regions; for each of the initially detected persons, extracting a first gradient feature and a second gradient feature of the person from the height-top-view, and a grayscale image or a color image corresponding to the height-top-view, respectively; for each of the initially detected persons, determining, based on the extracted first gradient feature and second gradient feature, using a previously constructed classifier corresponding to the determined visible region of the person, a confidence level of the initially detected person; and correcting, based on the confidence level, a detection result of the initially detected persons.

Abstract: A method and an apparatus for detecting an abnormal situation are disclosed. The method includes detecting whether a first target exists in an obtained image; recognizing whether the first target holds an object, when the first target exists in the image; obtaining motion information of the object, when the first target holds the object; and determining, based on the motion information of the object, whether the abnormal situation exists.

Abstract: A method and an apparatus for detecting an interest degree of a crowd in a target position are disclosed. The interest degree detection method includes projecting a depth image obtained by photographing onto a height-top-view, the depth image including the crowd and the target position; dividing the height-top-view into cells; determining density of the crowd in each cell; determining a moving speed and a moving direction of the crowd in each cell; determining orientation of the crowd in each cell; and determining, based on the density, the moving speed, the moving direction and the orientation of the crowd, the interest degree of the crowd in each cell in the target position. According to this method, the interest degree of the crowd in the target position can be detected accurately, even at a crowded place where it is difficult to detect and track a single person.

Abstract: Disclosed is an object dangerousness recognition method comprising steps of generating, based on an image captured by a stereo camera, a heterogeneous point cloud of an object in the image, each point in the heterogeneous point cloud having depth information and planar image information; determining, based on the depth information and the planar image information of each point in the heterogeneous point cloud, a solid shape of the object, and then, generating a first dangerousness parameter according to the solid shape; determining, based on the depth information and the planar image information of each point in the heterogeneous point cloud, a surface feature of the object, and then, generating a second dangerousness parameter according to the surface feature; and generating, based on the first and second dangerousness parameters, a comprehensive dangerousness parameter of the object.

Abstract: A method and a system for updating a background model based on depth are disclosed. The method includes receiving, in response to the occurrence of a predetermined background updating condition, one or more depth images captured after a time when the predetermined background updating condition occurs; obtaining, based on an original background model, foreground images in the one or more captured depth images, which are newly added compared with a depth image at the time when the predetermined background updating condition occurs; for each of foreground pixels in each of the newly added foreground images, comparing a current depth value with a previous depth value before the time when the predetermined background updating condition occurs; and updating, when the current depth value is greater than the previous depth value, the original background model as the updated background model by using the foreground pixel in the newly added foreground image.

Abstract: Disclosed are a shadow detection method and device. The method includes a step of obtaining a depth/disparity map and color/grayscale image from a two-lens camera or stereo camera; a step of detecting and acquiring plural foreground points; a step of projecting the acquired plural foreground points into a 3-dimensional coordinate system; a step of carrying out, in the 3-dimensional coordinate system, a clustering process with respect to the projected plural foreground points so as to divide the projected plural foreground points into one or more point clouds; a step of calculating density distribution of each of the one or more point clouds by adopting a principal component analysis algorithm so as to obtain one or more principal component values of the corresponding point cloud; and a step of determining, based on the one or more principal component values, whether the corresponding point cloud is a shadow.

Abstract: Disclosed are a scene video switch system and a scene video switch method. The system comprises an input unit configured to input an image currently captured by a video capture unit to serve as a current image; a detection unit configured to detect all of interesting elements and latent elements in the current image so as to determine a maximum possible existing area of the interesting elements; a state switch unit configured to carry out, by using a predetermined finite state machine, switching between scene states of the maximum possible existing area according to predetermined different scene states and switching conditions; and a display unit configured to extract, based on a current scene state switched by the state switch unit, an area of concern coinciding with the current scene state from the maximum possible existing area to serve as a scene video for display.