Research On Image Sub-pixel Detection Based On Wavelet Transform

With good time-frequency properties and multi-resolution characteristics, wavelet has a broad application prospect in image processing area. The paper researches image filtering method and edge detection with wavelet technology on the basis of classical theories,and develops sub-pixel image detection technology . The principle work of the paper includes:(1) Image filtering technology is mainly expounded. And review of the fundamental theory of edge detection is given.(2) Theories of wavelet transform are summarized. Principles of sub-pixel edge location are discussed, as well as the current sub-pixel edge location methods are developed. Then a least-square fitting method for sub-pixel edge location is proposed emphatically.(3) An algorithm for piston rings surface defects detection based on wavelet filtering and edge detection is developed, in which not only surface defects detection models are built, but also piston rings width measurement method is introduced in detail.(4) An algorithm, based on wavelet, for inspecting and measuring screw thread is reviewed. In engineering application, some feasible methods present in the paper for inspecting screw thread.In conclusion, this paper provides some methods related with wavelet and sub-pixel technology in applications, which adequate to solve problems of online machine-vision inspection in manufacturing.