Abstract: Embodiments of the present invention provide a method and a system for analyzing and learning behavior based on an acquired stream of video frames. Objects depicted in the stream are determined based on an analysis of the video frames. Each object may have a corresponding search model used to track an object's motion frame-to-frame. Classes of the objects are determined and semantic representations of the objects are generated. The semantic representations are used to determine objects' behaviors and to learn about behaviors occurring in an environment depicted by the acquired video streams. This way, the system learns rapidly and in real-time normal and abnormal behaviors for any environment by analyzing movements or activities or absence of such in the environment and identifies and predicts abnormal and suspicious behavior based on what has been learned.

Abstract: Techniques are described for detecting anomalous events using a long-term memory in a video analysis system. The long-term memory may be used to store and retrieve information learned while a video analysis system observes a stream of video frames depicting a given scene. Further, the long-term memory may be configured to detect the occurrence of anomalous events, relative to observations of other events that have occurred in the scene over time. A distance measure may used to determine a distance between an active percept (encoding an observed event depicted in the stream of video frames) and a retrieved percept (encoding a memory of previously observed events in the long-term memory). If the distance exceeds a specified threshold, the long-term memory may publish the occurrence of an anomalous event for review by users of the system.

Abstract: A long-term memory used to store and retrieve information learned while a video analysis system observes a stream of video frames is disclosed. The long-term memory provides a memory with a capacity that grows in size gracefully, as events are observed over time. Additionally, the long-term memory may encode events, represented by sub-graphs of a neural network. Further, rather than predefining a number of patterns recognized and manipulated by the long-term memory, embodiments of the invention provide a long-term memory where the size of a feature dimension (used to determine the similarity between different observed events) may grow dynamically as necessary, depending on the actual events observed in a sequence of video frames.

Abstract: Embodiments of the present invention provide a method and a module for identifying a background of a scene depicted in an acquired stream of video frames that may be used by a video-analysis system. For each pixel or block of pixels in an acquired video frame a comparison measure is determined. The comparison measure depends on difference of color values exhibited in the acquired video frame and in a background image respectively by the pixel or block of pixels and a corresponding pixel and block of pixels in the background image. To determine the comparison measure, the resulting difference is considered in relation to a range of possible color values. If the comparison measure is above a dynamically adjusted threshold, the pixel or the block of pixels is classified as a part of the background of the scene.

Abstract: Embodiments of the present invention provide a method and a module for identifying a background of a scene depicted in an acquired stream of video frames that may be used by a video-analysis system. For each pixel or block of pixels in an acquired video frame a comparison measure is determined. The comparison measure depends on difference of color values exhibited in the acquired video frame and in a background image respectively by the pixel or block of pixels and a corresponding pixel and block of pixels in the background image. To determine the comparison measure, the resulting difference is considered in relation to a range of possible color values. If the comparison measure is above a dynamically adjusted threshold, the pixel or the block of pixels is classified as a part of the background of the scene.

Abstract: Embodiments of the present invention provide a method and a module for identifying a background of a scene depicted in an acquired stream of video frames that may be used by a video-analysis system. For each pixel or block of pixels in an acquired video frame a comparison measure is determined. The comparison measure depends on difference of color values exhibited in the acquired video frame and in a background image respectively by the pixel or block of pixels and a corresponding pixel and block of pixels in the background image. To determine the comparison measure, the resulting difference is considered in relation to a range of possible color values. If the comparison measure is above a dynamically adjusted threshold, the pixel or the block of pixels is classified as a part of the background of the scene.

Abstract: A system is disclosed for facilitating speech recognition and transcription among users employing incompatible protocols for generating, transcribing, and exchanging speech. The system includes a system transaction manager that receives a speech information request from at least one of the users. The speech information request includes formatted spoken text generated using a first protocol. The system also includes a speech recognition and transcription engine, which communicates with the system transaction manager. The speech recognition and transcription engine receives the speech information request from the system transaction manager and generates a transcribed response, which includes a formatted transcription of the formatted speech. The system transmits the response to the system transaction manager, which routes the response to one or more of the users. The latter users employ a second protocol to handle the response, which may be the same as or different than the first protocol.

Abstract: A system is disclosed for facilitating speech recognition and transcription among users employing incompatible protocols for generating, transcribing, and exchanging speech. The system includes a system transaction manager that receives a speech information request from at least one of the users. The speech information request includes formatted spoken text generated using a first protocol. The system also includes a speech recognition and transcription engine, which communicates with the system transaction manager. The speech recognition and transcription engine receives the speech information request from the system transaction manager and generates a transcribed response, which includes a formatted transcription of the formatted speech. The system transmits the response to the system transaction manager, which routes the response to one or more of the users. The latter users employ a second protocol to handle the response, which may be the same as or different than the first protocol.

Abstract: A camera positioning assembly including a backing board, a base member connected to said backing board about an axis, a gear adapted for rotational movement, an arm connected to said gear and said base member for moving said base member about said axis in order that rotational movement of said gear is translated into rotational movement of said base member and a servo-controller that controls the rotational movement of said gear.

Abstract: An identification system is disclosed which determines whether the claimed identity of an individual is in fact the true identity. A three dimensional image of a hand is obtained using a digitizing camera and a pair of orthogonal reflecting surfaces. The image is then compared with a previously recorded hand image for the claimed identity. If the images match, within a selectable tolerance, the person is accepted as the correct individual. If they do not match, the person is rejected. Signalling and/or control apparatus responds to the results of the comparison.

Abstract: A robotic apparatus and method of acquiring jumbled objects or workpieces from a bin, and transferring them to a final site, or to an intermediate site for quality inspection. A video camera system produces a video signal representing a brightness image of the jumbled objects. An intermediate amplitude range of the video signal is selected, expanded, stored, and quantized into digital pixels. A computer enhances the image using histogram techniques, performs edge suppression, performs repeated shrinking with progressively-increased pixel thresholds until an area limit is reached, performs clustering of closely-spaced pixels, and commands movement of the robot gripper to a computed grip site on a recognized object. The gripper may limit its gripping pressure as a function of the deformation of the object gripped.

Abstract: A non-contacting optical probe incorporates an optical system designed to measure distances between various surfaces which may be internal or external such as exists in molds and the like. A collimated light beam is passed off-center through a first lens and focused to a first point on a first surface coplanar with the focal plane of the lens. When the distance between the first surface and the optical probe carrying the collimated beam changes so that a second surface is presented, this second surface is no longer in the focal plane with the result that a laterally displaced image of the point of interception of the collimated beam with the point of the intersection of the optical axis of the system occurs. This lateral displacement is imaged back through the optical system to a sensing surface.

Abstract: Disclosed herein is a system for selectively mordanting a dye to preselected areas of a black and white photographic silver image and rapidly removing all traces of the dye from the non-image areas. The system includes novel processing compositions to enable the entire process to be carried out rapidly with a minimum amount of equipment.

Abstract: A document sorting system employing a plurality of elongated flexible strips which form document transport channels therebetween and which terminate at a plurality of sorting areas. A magnetic apparatus is provided for selectively deflecting the ends of the flexible strips to form an appropriate channel entrance for the document to be sorted. A coil member is attached to the said end of each flexible strip and current is sent through the coil of the coil members. The coil members are disposed for movement on a piece of the magnetic apparatus and current through the coils of the coil members cooperate with the magnetic field produced by the magnetic apparatus to selectively deflect the coil members and the ends of the flexible strips. The coils of the coil members are connected in a circuit which determines the direction of the current flow in each of the coils.

Abstract: A reference diffraction pattern is generated by radiating an invariant physical characteristic such as a fingerprint with coherent light, this reference diffraction pattern being converted by a detector into a reference electrical function which is stored. A person to be identified then has a corresponding invariant physical characteristic such as his fingerprint radiated by coherent light to cause generation of a sample diffraction pattern. The frequency domain of the entire sample diffraction pattern is detected simultaneously to provide a sample electrical function. The reference electrical function in storage is then compared with the sample electrical function and if a correlation exists within a given tolerance, it is known that the person from which the sample electrical function is derived corresponds to the person from which the reference electrical function was derived. The invariant physical characteristics may be fingerprints as mentioned, or signatures, or combinations of both.

Abstract: Rather than counting nulls in the diffraction pattern generated by a very narrow line or gap in order to measure its width, the diffraction pattern is optically detected and converted to an analog electrical signal constituting a function of the detected intensities in the diffraction pattern. This function is then converted to a digital form, stored, and then passed to a computer electrically computing the Fourier transform to provide an autocorrelation wave form. This wave form is then plotted in an orthogonal X-Y coordinate system and the break point or extension of a linear segment of the autocorrelation curve relative to the X-axis used to provide a precise measurement of the given width.

Abstract: A system and method for character recognition in which each of a plurality of characters is recognized both magnetically and optically. If the magnetic reader fails to recognize a character while the optical reader is successful in recognizing it, a character identification signal corresponding to the character recognized by the optical reader is generated while if the optical reader fails to recognize a character and the magnetic reader is successful in recognizing it, an identification signal corresponding to the character recognized by the magnetic reader is generated. Where the magnetic and topical readers provide signals indicative of different characters while attempting to read the same character, depending on the parameters of the system an identification signal corresponding to one or the other recognition signals, or a reject signal is generated.