Geometric Distortion Correction Method Based On Anfis Plane Array Camera Study

In actual imaging of plane array camera, because of optical properties of lens,ect. different level of two-dimensional geometric distortion occur in image, it is bring adverse effects for useful of camera. With the plane array cameras are widely used in produce and research and so on, it is urgent to put adverse effects of geometric distortion to minimum, especially, in the field of high-precision measurement. So it is essential and based step for correcting geometric distortion of plane array camera.Geometric distortion of plane array camera cause pixel offset its ideal location in the image, so it need to find a mapping relationship for correcting geometric distortion of plane array camera, to make the actual pixels corresponds to ideal pixels. The mapping relationship can be a non-linear relationship. Adaptive Neuro-Fuzzy Inference System(ANFIS) can effectively carry out non-linear regression using method of Neural Networks and Fuzzy logic. The article introduce ANFIS in the correcting geometric distortion of plane array camera, and achieve the method by experiment. The article summarize the development of Neural Networks, Fuzzy system and methods of correct geometric distortion of plane array camera, analyze the non-linear imaging model of plane array camera. According to feature of target image, the article research method of extract intersection of the grid center in target image, and according to feature of distortion for plane array camera, calculate its ideal axis. It is achieve some useful methods and verify its accuracy. By the experimental comparison, verify the effectiveness of the correcting method of plane array camera based on ANFIS.