An Image Mosaic Method Based on Convolutional Neural Network Semantic Features Extraction

Since traditional image feature extraction methods rely on features such as corner points, a new method based on semantic feature extraction is proposed inspiring by convolution neural attack. The semantic features of each pixel in an image are computed and quantified by neural network to represent the contribution of each pixel to the image semantics. According to the quantization results, the semantic contribution values of each pixel are sorted, and the semantic feature points are selected from high to low and the image mosaic is completed. Experimental results show that this method can effectively extract image features and complete image mosaic.