A Binocular Stereo Matching Method Based on Joint Upsampling Convolutional Neural Network

Abstract

A binocular stereo matching method based on joint upsampling convolutional neural network, the method first uses the two-dimensional convolutional neural network based on joint upsampling to extract features from the input stereo image pair, and then combines the features of the stereo image pair The initial three-dimensional matching cost volume of the matching cost is constructed, and then three cascaded three-dimensional convolutional neural networks based on joint upsampling are used to aggregate the matching cost volume, and finally a dense disparity map with sub-pixel accuracy is obtained by using the regression method. Compared with the existing binocular stereo matching deep neural network, the present invention adopts a convolutional neural network based on pyramid joint upsampling in the decoding stage of feature extraction and cost aggregation steps, and by fusing multi-level and multi-scale contextual feature information, It can effectively retain more detailed textures in the upsampling process, and use depth-separable convolution with low computational complexity to improve the computational efficiency of the method and improve the quality of the disparity map for binocular stereo matching.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a binocular stereo matching method based on a joint upsampling convolutional neural network. Background technique [0002] Binocular stereo matching is a research problem that has attracted much attention in the field of computer vision. It has been widely used in various systems such as 3D reconstruction, automatic driving, autonomous robot navigation, and industrial inspection. In particular, current applications have an urgent need for real-time binocular stereo matching with high precision, high resolution, and large parallax. This undoubtedly poses a higher challenge to the computational efficiency and accuracy of the technology. In recent years, artificial intelligence based on deep learning technology has developed rapidly, and breakthroughs have been made in the fields of object detection, image classification, and speech recognition. Binocular stere...

AI Technical Summary

Problems solved by technology

Although the encoder can obtain rich semantic information by successively downsampling the spatial resolution, in the decoding process, the currently commonly used deconvolution upsampling tends to lose fine image structure information, resulting in parallax of object boundaries or small-sized objects Inaccurate prediction of outcome

Images