Abstract: A method and system for segmenting a plurality of images. The method comprises the steps of segmenting the image through a novel clustering technique that is, generating a composite depth map including temporally stable segments of the image as well as segments in subsequent images that have changed. These changes may be determined by determining one or more differences between the temporally stable depth map and segments included in one or more subsequent frames. Thereafter, the portions of the one or more subsequent frames that include segments including changes from their corresponding segments in the temporally stable depth map are processed and are combined with the segments from the temporally stable depth map to compute their associated disparities in one or more subsequent frames. The images may include a pair of stereo images acquired through a stereo camera system at a substantially similar time.

Abstract: A gesture control system includes a processor, the processor in communication with a plurality of sensors. The processor is configured to perform the steps of detecting, using the plurality of sensors, a gesture in a volume occupied by a plurality of occupants, analyzing a prior knowledge to associate the gesture with one of the plurality of occupants, and generating an output, the output being determined by the gesture and the one of the plurality of occupants.

Abstract: A method and system for globally updating a plurality of learning implicit gesture control systems. Embodiments can comprise receiving, by a global server and from a plurality of learning implicit gesture control systems, a user data. The global server configured to analyze the user data and determine an applicable integration level, and the global server further able to communicate with the plurality of learning implicit gesture control systems. Modifying a global parameter when the global server determines the applicable integration level is a global integration level; and transmitting the modified global parameter to the plurality of learning implicit gesture control systems.

Abstract: A method includes monitoring a plurality of system inputs, and detecting a behavioral pattern performed by a user and associated with the plurality of system inputs, When the behavioral pattern is detected, the method includes associating, in a memory, a gesture with at least one action, the at least one action being determined by the plurality of system inputs, and, upon detecting the gesture, executing the action associated with the gesture.

Abstract: A method and apparatus for creating an adaptive mosaic pixel-wise virtual Bayer pattern. The method may include receiving a plurality of monochromatic images from an array of imaging elements, creating a reference ordered set at infinity from the plurality of monochromatic images, running a demosaicing process on the reference ordered set, and creating a color image from the demosaiced ordered set. One or more offset artifacts resulting from the demosaicing process may be computed at a distance other than infinity, the ordered set may be modified in accordance with the computed offsets.

Abstract: A novel disparity computation technique is presented which comprises multiple orthogonal disparity maps, generated from approximately orthogonal decomposition feature spaces, collaboratively generating a composite disparity map. Using an approximately orthogonal feature set extracted from such feature spaces produces an approximately orthogonal set of disparity maps that can be composited together to produce a final disparity map. Various methods for dimensioning scenes and are presented. One approach extracts the top and bottom vertices of a cuboid, along with the set of lines, whose intersections define such points. It then defines a unique box from these two intersections as well as the associated lines. Orthographic projection is then attempted, to recenter the box perspective. This is followed by the extraction of the three-dimensional information that is associated with the box, and finally, the dimensions of the box are computed. The same concepts can apply to hallways, rooms, and any other object.

Abstract: A gesture-enabled keyboard and method are defined. The gesture-enabled keyboard includes a keyboard housing including one or more keyboard keys for typing and a pair of stereo camera sensors mounted within the keyboard housing, a field of view of the pair of stereo camera sensors projecting substantially perpendicularly to the plane of the keyboard housing. A background of the field of view is updated when one or more alternative input devices are in use. A gesture region including a plurality of interaction zones and a virtual membrane defining a region of transition from one of the plurality of interaction zones to another of the plurality of interaction zones is defined within the field of view of the pair of stereo camera sensors. Gesture interaction is enabled when one or more gesture objects are positioned within the gesture region, and when one or more alternative input devices are not in use.

Abstract: A gesture control system includes a processor, the processor in communication with a plurality of sensors. The processor is configured to perform the steps of detecting, using the plurality of sensors, a gesture in a volume occupied by a plurality of occupants, analyzing a prior knowledge to associate the gesture with one of the plurality of occupants, and generating an output, the output being determined by the gesture and the one of the plurality of occupants.

Abstract: A user, such as the driver of a vehicle, to retrieve information related to a point of interest (POI) near the vehicle by pointing at the POI or performing some other gesture to identify the POI. Gesture recognition is performed on the gesture to generate a target region that includes the POI that the user identified. After generating the target region, information about the POI can be retrieved by querying a server-based POI service with the target region or by searching in a micromap that is stored locally. The retrieved POI information can then be provided to the user via a display and/or speaker in the vehicle. This process beneficially allows a user to rapidly identify and retrieve information about a POI near the vehicle without having to navigate a user interface by manipulating a touchscreen or physical buttons.

Abstract: A near-touch interface is provided that utilizes stereo cameras and a series of targeted structured light tessellations, emanating from the screen as a light source and incident on objects in the field-of-view. After radial distortion from a series of wide-angle lenses is mitigated, a surface-based spatio-temporal stereo algorithm is utilized to estimate initial depth values. Once these values are calculated, a subsequent refinement step may be applied in which light source tessellations are used to flash a structure onto targeted components of the scene, where initial near-interaction disparity values have been calculated. The combination of a spherical stereo algorithm, and smoothing with structured light source tessellations, provides for a very reliable and fast near-field depth engine, and resolves issues that are associated with depth estimates for embedded solutions of this approach.

Abstract: A method and apparatus for segmenting an image are provided. The method may include the steps of clustering pixels from one of a plurality of images into one or more segments, determining one or more unstable segments changing by more than a predetermined threshold from a prior of the plurality of images, determining one or more segments transitioning from an unstable to a stable segment, determining depth for one or more of the one or more segments that have changed by more than the predetermined threshold, determining depth for one or more of the one or more transitioning segments, and combining the determined depth for the one or more unstable segments and the one or more transitioning segments with a predetermined depth of all segments changing less than the predetermined threshold from the prior of the plurality of images.

Abstract: A method and apparatus for processing image data is provided. The method includes the steps of employing a main processing network for classifying one or more features of the image data, employing a monitor processing network for determining one or more confusing classifications of the image data, and spawning a specialist processing network to process image data associated with the one or more confusing classifications.

Abstract: A novel disparity computation technique is presented which comprises multiple orthogonal disparity maps, generated from approximately orthogonal decomposition feature spaces, collaboratively generating a composite disparity map. Using an approximately orthogonal feature set extracted from such feature spaces produces an approximately orthogonal set of disparity maps that can be composited together to produce a final disparity map. Various methods for dimensioning scenes and are presented. One approach extracts the top and bottom vertices of a cuboid, along with the set of lines, whose intersections define such points. It then defines a unique box from these two intersections as well as the associated lines. Orthographic projection is then attempted, to recenter the box perspective. This is followed by the extraction of the three-dimensional information that is associated with the box, and finally, the dimensions of the box are computed. The same concepts can apply to hallways, rooms, and any other object.

Abstract: A system for capturing image data for gestures from a passenger or a driver in a vehicle with a dynamic illumination level comprises a low-lux sensor equipped to capture image data in an environment with an illumination level below an illumination threshold, a high-lux sensor equipped to capture image data in the environment with the illumination level above the illumination threshold, and an object recognition module for activating the sensors. The object recognition module determines the illumination level of the environment and activates the low-lux sensor if the illumination level is below the illumination threshold. If the illumination level is above the threshold, the object recognition module activates the high-lux sensor.

Abstract: A method and system to determine the disparity associated with one or more textured regions of a plurality of images is presented. The method comprises the steps of breaking up the texture into its color primitives, further segmenting the textured object into any number of objects comprising such primitives, and then calculating a disparity of these objects. The textured objects emerge in the disparity domain, after having their disparity calculated. Accordingly, the method is further comprised of defining one or more textured regions in a first of a plurality of images, determining a corresponding one or more textured regions in a second of the plurality of images, segmenting the textured regions into their color primitives, and calculating a disparity between the first and second of the plurality of images in accordance with the segmented color primitives.

Abstract: A method and apparatus for segmenting an image are provided. An image acquisition and segmentation apparatus may include a sensor, a lens having a variable distance to the sensor, and a processor for acquiring an image and segmenting one or more segments from the acquired image. The lens may be moved relative to the sensor in accordance with a microelectronic machine in order to provide a shallow depth of field and a resulting blurred background, therefore easing a segmentation of one or more foreground objects in an acquired image.

Abstract: A near-touch interface is provided that utilizes stereo cameras and a series of targeted structured light tessellations, emanating from the screen as a light source and incident on objects in the field-of-view. After radial distortion from a series of wide-angle lenses is mitigated, a surface-based spatio-temporal stereo algorithm is utilized to estimate initial depth values. Once these values are calculated, a subsequent refinement step may be applied in which light source tessellations are used to flash a structure onto targeted components of the scene, where initial near-interaction disparity values have been calculated. The combination of a spherical stereo algorithm, and smoothing with structured light source tessellations, provides for a very reliable and fast near-field depth engine, and resolves issues that are associated with depth estimates for embedded solutions of this approach.

Abstract: A method and apparatus for creating an adaptive mosaic pixel-wise virtual Bayer pattern. The method may include receiving a plurality of monochromatic images from an array of imaging elements, creating a reference ordered set at infinity from the plurality of monochromatic images, running a demosaicing process on the reference ordered set, and creating a color image from the demosaiced ordered set. One or more offset artifacts resulting from the demosaicing process may be computed at a distance other than infinity, the ordered set may be modified in accordance with the computed offsets.

Abstract: A user, such as the driver of a vehicle, to retrieve information related to a point of interest (POI) near the vehicle by pointing at the POI or performing some other gesture to identify the POI. Gesture recognition is performed on the gesture to generate a target region that includes the POI that the user identified. After generating the target region, information about the POI can be retrieved by querying a server-based POI service with the target region or by searching in a micromap that is stored locally. The retrieved POI information can then be provided to the user via a display and/or speaker in the vehicle. This process beneficially allows a user to rapidly identify and retrieve information about a POI near the vehicle without having to navigate a user interface by manipulating a touchscreen or physical buttons.

Abstract: A system for capturing image data for gestures from a passenger or a driver in a vehicle with a dynamic illumination level comprises a low-lux sensor equipped to capture image data in an environment with an illumination level below an illumination threshold, a high-lux sensor equipped to capture image data in the environment with the illumination level above the illumination threshold, and an object recognition module for activating the sensors. The object recognition module determines the illumination level of the environment and activates the low-lux sensor if the illumination level is below the illumination threshold. If the illumination level is above the threshold, the object recognition module activates the high-lux sensor.