5 results about "Rgb image" patented technology

The invention provides a behavior recognition method and system based on an attention mechanism double-flow network, and belongs to the technical field of behavior recognition, and the method comprises the steps: dividing an obtained whole video segment into a plurality of video segments with the same length, extracting an RGB image and an optical flow gray-scale image of each frame of each videosegment, and carrying out the preprocessing of the RGB images and the optical flow gray-scale images; carrying out random sampling on the preprocessed image to obtain an RGB image and an optical flowgrayscale image of each video clip; extracting appearance features and time dynamic features of the sampled images by using a double-flow network model introducing an attention mechanism, fusing the appearance features and the time dynamic features according to the types of a time domain network and a space domain network respectively, and performing weighted fusion on a fusion result of the timedomain network and a fusion result of the space domain network to obtain an identification result of the whole video. According to the invention, the video data can be fully utilized, the local key features of the video frame image can be better extracted, the foreground area where the action occurs is highlighted, the influence of irrelevant information in the background environment is inhibited,and the behavior recognition accuracy is improved.

The embodiment of the invention provides a demosaicing method and device, electronic equipment and a storage medium. The method comprises: obtaining a target RGBIR image; performing first processing on the target RGBIR image by using a first neural network to obtain a target IR image; and/or performing second processing on the target RGBIR image by using a second neural network to obtain a target RGB image, wherein the first processing comprises: eliminating a visible light component in a pixel value corresponding to an IR channel pixel point in the target RGBIR image and predicting a pixel value of a color channel pixel point in an IR channel, and the second processing comprises: eliminating a near-infrared light component in a pixel value corresponding to the color channel pixel point in the target RGBIR image for each color channel, and predicting the pixel values of the pixel points except the pixel points of the color channel in the color channel.

The invention provides a depth map reconstruction method combining laser points and mean shift. The depth map reconstruction method comprises the following implementation steps: (1) shooting a depth image; (2) shooting an RGB image containing laser points; (3) establishing a mean shift descriptor by combining a laser point track; (4) setting a weight according to a laser point rule; (5) setting athreshold value to judge holes and noise of the depth map; (6) de-noising the depth map and filling holes; And (7) reconstructing by combining the depth map and the RGB image. According to the method,a large number of object characteristics are reserved, high robustness is achieved, and a good reconstruction effect can be kept in various environments and on a shot object.

A system and method may include capturing a multi-channel polarimetric image and a multi-channel RGB image of a scene by a color polarimetric imaging camera. A multi-channel hyperspectral image may be synthesized from the multi-channel RGB image and concatenated with the multi-channel polarimetric image to create an integrated polarimetric-hyperspectral image. Scene properties within the integrated polarimetric-hyperspectral image may be disentangled.