Image data reduction with splines and segmentation with emphasis on earth observation

TL;DRAbstract

This report describes the results of a feasibility study on image data compression using spline approximation and image segmentation, with emphasis on earth observation applications. The main subjects dealt with, are: the data compression method, implementation aspects, the test software and test results. It appears that a reduction factor of 5 to 10 is possible without significant visual loss of information. The spline approximation involves a convolution process. A simple hardware implementation is described. Benchmark tests yielded an accuracy of 5 bits. Possible improvements are discussed.

Chat with Paper

AI Agents for this Paper