Wavelet and nonlinear model based borescope image compression
Ding Peng1, Ma Qishuang1, Li Changyou1, Yao Hongyu2*
1. School of Automation Science and Electrical Engineering, Beijing University of Aeronautics and Astronautics, Beijing 100083, China;
2. Center of Aviation Safety Technology of China, General Administration of Civil Aviation of China, Beijing 100028, China
The borescope image quality around edges is very important to the fault diagnostic of aeronautic engine. And inside the turbofan engine, the borescope image often has complex edges and textures. To achieve high compression ratio without losing of the edge and texture information, a kind of image compression method with emphasis on edges and textures was presented. First, it used wavelet transform in order to obtain a set of biorthogonal subclasses of images. Second, multiscale Canny edge detector was used to locate the edge. Third, the AOI(area of influence) was set by contour plus profile model, witch was an edge nonlinear represent and modeling method. Finally, by rate control, different image quality was obtained in different areas of the image, like ROI(region-of-interest) coding of JPEG2000. Test result shows, comparing with general image compression method, at different compression ratio, PSNR(peak signal to noise ratio) has all gotten lightweight raise, and the visual perception around edges is also better than the general method.