Method of parameters identification of imaging model used to correct computer vision error

An imaging channel's nonlinear characteristic is one of the factors of computer vision error. Common methods of distortion correction rely on a lot of prior knowledge and interactive information. As the imaging model is only related to the structure of the vision system, and the parameters of the model can be identified automatically, it has a prevalent significance to correct computer vision error using an imaging model. The paper gives an example of a micro robot soccer vision system, using an imaging model, identifying parameters and gaining good results.