Abstract: A stereoscopic video generation method based on 3D convolution neural network is disclosed, which is able to convert existing 2D video sources into stereoscopic videos. The method includes preparing the training data, dividing the training video sources into left eye view sequences and right eye view sequences; and then processing the left eye image sequences through shot segmentation via fuzzy C-means clustering method, calculating a mean image of all left eye images, subtracting the mean image from the left eye images, taking the right eye images as a training target; training the obtained 3D convolution neural network through the training data; processing the 2D video sources which need to be converted into stereoscopic videos in the same way as training set, inputting to the trained 3D convolution neural network to obtain the right eye view image sequences of the 2D videos; and finally combining the two to be stereoscopic videos.

Abstract: In some embodiments, systems, apparatuses, and methods are described herein including a shelf assembly that utilizes the accumulated weight of products stocked thereon to compress a compression member disposed within the assembly. An electronic imaging device can be mounted proximate to the assembly and can be oriented to capture an image of the compression member to determine whether a particular shelf needs to be restocked and avoid having to manipulate or remove products from a shelf to determine a current stock level.

Abstract: A light source illumination unit has operation modes including at least a normal electric power mode to operate with first electric power consumption, and a low electric power mode to operate with second electric power consumption lower than the first electric power consumption. A controller selects at least one of the operation modes in accordance with a use schedule, and controls the light source illumination unit in accordance with the selected operation mode.

Abstract: A noise data update processing unit calculates the amount of a fixed pattern noise component on the basis of a detection signal of infrared rays detected by an infrared detector in a state where an optical system is controlled to be in a non-focused state, and updates an FPN data storage unit with the calculated amount of the FPN component. The noise data update processing unit calculates an average value of detection signals of each detector element and a plurality of peripheral detector elements, and calculates a signal component dependent on incident infrared rays included in the detection signal of each detector element by subtracting an average value of fixed pattern noise data before update from the calculated average value. The amount of the fixed pattern noise component is calculated by subtracting the calculated signal component from the detection signal of each detector element.

Abstract: An image pickup system includes: an image pickup section including a first image pickup device outputting a first image pickup signal, and a second image pickup device outputting a second image pickup signal; a processor that performs signal processing on image pickup signals; and a memory section holding a difference correction parameter, the difference correction parameter indicating difference of image characteristics derived from sensitivity of each image pickup device, in image pickup signals outputted from the first image pickup device and the second image pickup device; and a correction processing section performing correction processing on at least one of the first image pickup signal and the second image pickup signal, based on the difference correction parameter.

Abstract: In a digital contents receiver for receiving transmitted digital contents, the digital contents include at least component information indicating an element which constitutes a program of the contents. When the component information indicates that the received digital contents are a 3D component, it is determined whether a display part corresponds to display of the 3D component. If the display part corresponds to display of the 3D component, the received digital contents are displayed in 3D.

Abstract: A rearview mirror for a vehicle includes a housing defining an interior cavity and an open side. A substrate is rigidly coupled within the open side of the housing and has a reflective surface thereon. The rearview mirror further includes an actuation mechanism having a mounting plate rotatably coupled within the cavity of the housing at a first end of the mounting plate. A motor is rigidly coupled within the interior cavity of the housing adjacent a second end of the mounting plate. The motor has a first wheel coupled with an output shaft thereof, and the first wheel includes a pin coupled therewith that extends outwardly from the pin at a location offset from the output shaft of the motor. A first linkage arm is rotatably coupled between the pin of the first wheel and the second end of the mounting plate.

Abstract: A system and method for focusing a camera is disclosed. The system may generate a proxy image that is blurred to an extent that is correlated to a degree to which a camera is out of focus. A user may be asked to adjust a focusing mechanism to attempt to bring the proxy image into focus. This allows the camera to be focused without the user needing to see an image from the camera. This can be used, for example, to focus an infrared camera. The infrared camera could be a tracking camera in a device such as a head mounted display device.

Abstract: A display device includes a display unit that displays an image including a stereoscopic image based on binocular parallax, a detection unit that detects an obstacle that blocks the stereoscopic image, and a control unit that determines at least one of a display state and a blocking state of the stereoscopic image by the obstacle based on a detection result of the detection unit, and controls a display of a related image based on the determination result, the related image being related to the stereoscopic image.

Abstract: An imaging system is disclosed. The system comprises: a first imaging device and a second imaging device being spaced apart and configured to provide partially overlapping field-of-views of a scene over a spectral range from infrared to visible light. The system comprises at least one infrared light source constituted for illuminating at least the overlap with patterned infrared light, and a computer system configured for receiving image data pertaining to infrared and visible light acquired by the imaging devices, and computing three-dimensional information of the scene based on the image data. The image data optionally and preferably comprises the patterned infrared light as acquired by both the imaging devices.

Abstract: Disclosed herein is a method of encoding a video signal, comprising receiving an original video signal; comparing the original video signal with available reconstructed signals; determining a correction signal based on a result of the comparison; and generating a transform-coded correction signal to be transmitted for a signal reconstruction.

Abstract: An optical system for an image acquisition device comprises an image sensor comprising an array of pixels including pixels sensitive to IR wavelengths for acquiring an image. A lens assembly includes a collecting lens surface with an optical axis, the lens assembly being arranged to focus IR light received from a given object distance on the sensor surface. The lens assembly includes at least a first reflective surface for reflecting collected light along an axis transverse to the optical axis so that a length of the optical system along the optical axis is reduced by comparison to a focal length of the lens assembly.

Abstract: Systems, apparatuses and methods may provide for detecting a snapshot request to conduct a long range depth capture, wherein the snapshot request is associated with a short range depth capture. Additionally, an infrared (IR) projector may be activated at a first power level for a first duration in response to the snapshot request, wherein the first power level is greater than a second power level corresponding to the short range depth capture and the first duration is less than a second duration corresponding to the short range depth capture.

Abstract: A signal component amount calculation unit calculates dispersion or standard deviation of multiple times of infrared detection signals detected by each detector element to be processed, and calculates the amount of a signal component dependent on infrared rays incident on the infrared detector included in the infrared detection signals, on the basis of the calculated dispersion or standard deviation. A fixed pattern noise calculation unit calculates the amount of a fixed pattern noise component on the basis of the infrared detection signals and the calculated amount of a signal component. A data update unit updates the fixed pattern noise data with the calculated amount of a fixed pattern noise component.

Abstract: A vehicle authentication system may include a controller configured to detect a trigger event, activate at least one camera, transmit a unique identifier to a mobile device, receive an image from the at least one camera, and grant access to the vehicle in response to the image including the unique identifier.

Abstract: Provided are an infrared imaging device and a signal correction method using the infrared imaging device capable of reducing noise due to a variation of pixel signals of an infrared image even in a case where the variation of the pixel signals of the infrared image is changed. An infrared imaging device includes an optical system, a detection unit that includes an effective pixel portion in which effective pixels that are infrared detection elements are arranged in a matrix form in an effective region onto which infrared rays from the optical system are incident and a reference pixel portion in which a plurality of reference pixels that are the infrared detection elements are arranged in a reference region onto which the infrared rays from the optical system are not incident, and is positioned on an imaging surface of the optical system; and a signal correction unit that performs a correction process with respect to a pixel signal detected by the detection unit.

Abstract: The present invention provides an infrared image reconstruction method based on block-sparse compressive sensing. The method includes: a block-sparse representation step: representing consecutive infrared small target image sequences as signals having block-sparse characteristics; a temporal correlation modeling step: modeling a temporal correlation of data between the consecutive infrared small target image sequences under a Bayesian framework; and a signal reconstruction and output step: reconstructing infrared small target images using a sparse learning algorithm and outputting final estimates of the signals. The present invention further provides an infrared image reconstruction system based on block-sparse compressive sensing.

Abstract: Related to the field of security protection, a video surveillance method, apparatus, and system, which can implement local magnification of a moving object in a surveillance scene. The method includes obtaining, by a client, an intelligent analysis alarm and moving track information of at least one alarm triggering object, and adjusting a window layout of a user interface according to the intelligent analysis alarm, where the intelligent analysis alarm is triggered by the client, a surveillance platform, or a camera when determining that there is an intrusion object in a surveillance scene according to an intelligent analysis strategy; and performing, by the client, local magnification on the at least one alarm triggering object according to moving track information of the at least one alarm triggering object. The present disclosure is applicable to video surveillance of important places.

Abstract: A three-dimensional (3D) video encoding method, decoding method, and related apparatus is disclosed. The decoding method may include decoding a video bitstream to obtain a single sample flag bit corresponding to a current image block in a current depth map, performing detection on a first adjacent prediction sampling point and a second adjacent prediction sampling point of the current image block in the current depth map if the single sample flag bit obtained by decoding indicates that a decoding mode corresponding to the current image block is a single depth intra-frame mode (SDM), and constructing a sample candidate set according to results of the detection on the first adjacent prediction sampling point and the second adjacent prediction sampling point, where the sample candidate set includes a first index location and a second index location, decoding the video bitstream to obtain a single sample index flag bit corresponding to the current image block.

Abstract: A method and an apparatus for detecting a pedestrian by a vehicle during night driving are provided, in which the apparatus includes: a first camera configured to take a first image including color information of a vicinity of the vehicle during night driving; a second camera configured to take a second image including thermal distribution information of the vicinity of the vehicle; a pedestrian detector configured to detect a non-pedestrian area by using the color information from the first image and detect a pedestrian area by excluding the non-pedestrian area from the second image; and a display configured to match and display the pedestrian area on the second image.