Research On Compression Technology Of Photogrammetric Image

The popularity of high-resolution industrial cameras provides a severe challengeto the transmission and storage of photogrammetric images, because thesephotogrammetric images contain a huge mass of data. Thereforce, it has a greatapplication value to design a compression scheme that is sutitable forphotogrammetric images. In other words, it is necessary to encode thephotogrammetric images to save the storage space.According to the characteristics of photogrammetric images and the correlativetheory of image compression, the compression technology of high-resolutionphotogrammetric image is researched in detail. In general, the major contents andinnovations achieved in this paper are shown as following:1. The image compression methods used in the common vision measurementsystems are discussed. According to the calculation methods of vision measurementtechnology, the compression requirement and the encoding implement scheme ofphotogrammetric images are proposed by combining the application features ofphotogrammetric image with the image encoding technology. After summarizingthe research status of ROI encoding theory, the technical details and the evaluatingindicators of the encoding scheme, used for encoding the photogrammetric images,are described in detail.2. The extraction methods of feature regions in photogrammetric images areresearched in detail. After analyzing the extraction methods of feature regions in laserfacula images, elliptical target images and structured light strrpe images, the generalextraction method of feature regions in photogrammetric images is proposed usingimage binarization and the image dilation algorithm. In order to reduce the repeatedcalculations existing in the image dilation, a modified image dilation method based onthe Freeman chain code is provided. And the general extraction method of featureregions in photogrammetric images is verified by the experiment.3. The encoding scheme of photogrammetric images based on bitplane lifting isstudied. In this scheme, the encoding of arbitrary resolution photogrammetric image isachieved by combining the regions coding and modified SPIHT coder. In themodified SPIHT coder, the encoding efficiency of photogrammetric images isimproved by the sperate encoding between the sign bitplane and the magnitudebitplanes. A quantization scheme based on high frequency wavelet subband optimizesthe output bit stream in the sorting pass. According to the energy concentration features of wavelet transform, only the back ground coefficients belonging to thelowest frequency sub-band are encoded in order to improve total compression ratio.The verification experiment is achieved to test the feasibility of this encoding scheme.4. The encoding scheme of photogrammetric images based on ROI mask isproposed. According to the distributive characters of ROI mask inphotogrammetric images, two encoding methods of ROI mask are proposed to reducethe bitrate. The compression performance of photogrammetric images is greatlyimproved using the modified partitioning structure, the improved encoding ofmagnitude bitplanes and the PROI mask. Only the coefficients indicated by ROI maskin the low frequency subband are processed to improve total compressionperformance. Finally, the coding performance for photogrammetric images is verifiedby the experiment.5. In the large industrial measurement field, the application of the encodingscheme based on ROI mask is researched in detail. The compression performancesand the measurement precisions are discussed and analyied in detail in the differentthreshods and the different radiuses of the expansion elements. The experiment resultsshow that the scheme has higher compression performance forphotogrammetric images without affecting the measurement precision.