Research on multi-camera information fusion method for intelligent perception

In this paper, the Gaussian Mixture Model and Mean Shift algorithm are used to detect and track moving objects in the visual perception network composed of multiple cameras. And on this basis, a target matching method based on wavelet transform, which is applied in a visual perception network composed by multiple camera, fusing visual information from different cameras is proposed. This method takes local features as basis of target matching, and applies wavelet transformation to detect the feature points that represent important information of the target image, and then extracts the color of the neighborhood of feature points as its salient features. The method of classification and clustering is applied by calculating the distance of salient features vector space to measure similarities of the target features and thus realize target recognition. The test result shows that the method can realize the matching and recognition of moving object with the cooperation among multiple cameras.