Abstract: Systems and methods for robotic detection of escalators are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may navigate a learned route and utilize one or more methods of detecting an escalator using data from its sensors. The robot may subsequently avoid the area comprising the escalator.

Abstract: Systems and methods for detecting floor from noisy depth measurements for robots are disclosed herein. According to at least one non-limiting exemplary embodiment, a height map may be produced based on one or more depth measurements from a sensor of a robot. The height map may be utilized to determine surface normal vectors which may be further utilized by the robot to determine if regions of the height map are floor.

Abstract: A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.

Abstract: Systems, apparatuses, and methods for calibrating LiDAR sensors of a robot using intersecting LiDAR sensors are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may calibrate a calibration LiDAR based on a determined pose of the calibration LiDAR, wherein the pose is determined based on a measurement error between the calibration LiDAR and an intersecting reference LiDAR.

Abstract: Systems, apparatuses, and methods for calibrating LiDAR sensors of a robot using intersecting LiDAR sensors are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may calibrate a calibration LiDAR based on a determined pose of the calibration LiDAR, wherein the pose is determined based on a measurement error between the calibration LiDAR and an intersecting reference LiDAR.

Abstract: A target image captured from a fisheye lens or other lens with known distortion parameters may be transformed to align it to a reference image. Corresponding features may be detected in the target image and the reference image. The features may be transformed to a spherical coordinate space. In the spherical space, images may be re-pointed or rotated in three dimensions to align all or a subset of the features of the target image to the corresponding features of the reference image. For example, in a sequence of images, background features of the target image in the spherical image space may be aligned to background features of the reference image in the spherical image space to compensate for camera motion while preserving foreground motion. An inverse transformation may then be applied to bring the images back into the original image space.

Abstract: A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.

Abstract: Methods and apparatus for the generation of interpolated frames of video data. In one embodiment, the interpolated frames of video data are generated by obtaining two or more frames of video data from a video sequence; determining frame errors for the obtained two or more frames from the video sequence, determining whether the frame errors exceed a threshold value; performing a multi-pass operation; performing a single-pass operation; performing frame blending; performing edge correction; and generating the interpolated frame of image data.

Abstract: A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.

Abstract: An algorithm for performing an image or video processing task is generated that may be used to combine a plurality of different independent solutions to the image or video processing task in an optimized manner. A plurality of base algorithms may be applied to a training set of images or video and a first generation of different combining algorithms may be applied to combine the respective solutions from each of the respective base algorithms into respective combined solutions. The respective combined solutions may be evaluated to generate respective fitness scores representing measures of how well the plurality of different combining algorithms each perform the image or video processing task. The algorithms may be iteratively updated to generate an optimized combining algorithm that may be applied to an input image or video.

Abstract: A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.

Abstract: Methods and apparatus for the generation of interpolated frames of video data. In one embodiment, the interpolated frames of video data are generated by obtaining two or more frames of video data from a video sequence; determining frame errors for the obtained two or more frames from the video sequence, determining whether the frame errors exceed a threshold value; performing a multi-pass operation; performing a single-pass operation; performing frame blending; performing edge correction; and generating the interpolated frame of image data.

Abstract: Video and corresponding metadata is accessed. Events of interest within the video are identified based on the corresponding metadata, and best scenes are identified based on the identified events of interest. In one example, best scenes are identified based on the motion values associated with frames or portions of a frame of a video. Motion values are determined for each frame and portions of the video including frames with the most motion are identified as best scenes. Best scenes may also be identified based on the motion profile of a video. The motion profile of a video is a measure of global or local motion within frames throughout the video. For example, best scenes are identified from portion of the video including steady global motion. A video summary can be generated including one or more of the identified best scenes.

Abstract: A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.

Abstract: A target image captured from a fisheye lens or other lens with known distortion parameters may be transformed to align it to a reference image. Corresponding features may be detected in the target image and the reference image. The features may be transformed to a spherical coordinate space. In the spherical space, images may be re-pointed or rotated in three dimensions to align all or a subset of the features of the target image to the corresponding features of the reference image. For example, in a sequence of images, background features of the target image in the spherical image space may be aligned to background features of the reference image in the spherical image space to compensate for camera motion while preserving foreground motion. An inverse transformation may then be applied to bring the images back into the original image space.

Abstract: Methods and apparatus for the generation of interpolated frames of video data. In one embodiment, the interpolated frames of video data are generated by obtaining two or more frames of video data from a video sequence; determining frame errors for the obtained two or more frames from the video sequence, determining whether the frame errors exceed a threshold value; performing a multi-pass operation; performing a single-pass operation; performing frame blending; performing edge correction; and generating the interpolated frame of image data.

Abstract: A target image captured from a fisheye lens or other lens with known distortion parameters may be transformed to align it to a reference image. Corresponding features may be detected in the target image and the reference image. The features may be transformed to a spherical coordinate space. In the spherical space, images may be re-pointed or rotated in three dimensions to align all or a subset of the features of the target image to the corresponding features of the reference image. For example, in a sequence of images, background features of the target image in the spherical image space may be aligned to background features of the reference image in the spherical image space to compensate for camera motion while preserving foreground motion. An inverse transformation may then be applied to bring the images back into the original image space.

Abstract: Video and corresponding metadata is accessed. Events of interest within the video are identified based on the corresponding metadata, and best scenes are identified based on the identified events of interest. In one example, best scenes are identified based on the motion values associated with frames or portions of a frame of a video. Motion values are determined for each frame and portions of the video including frames with the most motion are identified as best scenes. Best scenes may also be identified based on the motion profile of a video. The motion profile of a video is a measure of global or local motion within frames throughout the video. For example, best scenes are identified from portion of the video including steady global motion. A video summary can be generated including one or more of the identified best scenes.

Abstract: Video and corresponding metadata is accessed. Events of interest within the video are identified based on the corresponding metadata, and best scenes are identified based on the identified events of interest. In one example, best scenes are identified based on the motion values associated with frames or portions of a frame of a video. Motion values are determined for each frame and portions of the video including frames with the most motion are identified as best scenes. Best scenes may also be identified based on the motion profile of a video. The motion profile of a video is a measure of global or local motion within frames throughout the video. For example, best scenes are identified from portion of the video including steady global motion. A video summary can be generated including one or more of the identified best scenes.

Abstract: Apparatus and methods for video image post-processing for segmentation based interpolation. In one embodiment, a computerized apparatus is utilized in order to obtain a first frame of video data; segment one or more objects within the first frame of video data; obtain a second frame of video data; segment one or more objects within the second frame of video data; match at least a portion of the one or more objects within the first frame of video data with the one or more objects within the second frame of video data; compute the motion of the pixels for the matched portion of the one or more objects; compute the motion of pixels associated with a background image; and generate an interpolated frame of video data, the interpolated frame of video data residing temporally between the first frame of video data and the second frame of video data.