IQ evaluation based adaptive wavelet denoising and enhancement for a VTRAN system

An Image Quality (IQ) evaluation based wavelet domain denoising and enhancement model for a Vehicle Target Recognition and Assistant Navigation (VTRAN) system is introduced. In order to adapt to a complex atmosphere environment, our model utilizes the IQ parameters as a criterion to search the proper estimation values of the wavelet domain denoising and enhancement model. Firstly, before the wavelet processing, we evaluate the IQ and estimate the expected image gray values from some sub-regions of the original image. Then our algorithm will add some ldquocleanrdquo and ldquoclearrdquo points with those expected gray values in the original image. After that our model will calculate the wavelet based denoising and enhancement model and evaluate the result by the IQ with those added reference points. Finally, if the IQ is not good enough, a feedback will be given and the model parameters will be modified until it gets a better result. Experiment results show our algorithm works well in many outdoor tests and it can be used in a complex atmosphere environment.