Based On DIBR Compression Technology Research

Integrate image technology is one of the most important research direction in current thr ee-dimensional image technology field. It can be observed through the three-dimensional images with a real depth scene. However, the payment of this increasing of visual realism is the volume of the data of single-channel will be larger in times. Therefore, the3D image compression encoding method is one of the most hotspot in3D image research areas.In this paper, we start from the storage of the elemental image sequence that collects from integrate imaging technology, and the image redundancy between integrated image row/column is very high, and we using the method of extract sampling of row/column of images and then to remove redundant information in order to compress, integrate image compression algorithm is to be proposed that based on DIBR and the combination of neural network, and that can accurately restore the image position and color of the image which are be extracted. The experiment shows that the algorithm proposed in this paper has better results. Integrate image after the processing of compression, and then be decompressed, the result of decompressed image quality are higher than that being restored by JPEG, and it also be better in other elements that compared to other Integrate image compression.