Quality Evaluation of Adaptive Optical Image Based on DCT and Renyi Entropy

The adaptive optical telescopes play a more and more important role in the detection system on the ground, and the adaptive optical images are so many that we need find a suitable method of quality evaluation to choose good quality images automatically in order to save human power. It is well known that the adaptive optical images are no-referrence images. In this paper, a new logarithmic evaluation method based on the use of the discrete cosine transform(DCT) and Renyi entropy for the adaptive optical images is proposed. Through the DCT using one or two dimension window, the statistical property of Renyi entropy for images is studied. The different directional Renyi entropy maps of an input image containing different information content are obtained. The mean values of different directional Renyi entropy maps are calculated. For image quality evaluation, the different directional Renyi entropy and its standard deviation corresponding to region of interest is selected as an indicator for the anisotropy of the images. The standard deviation of different directional Renyi entropy is obtained as the quality evaluation value for adaptive optical image. Experimental results show the proposed method that the sorting quality matches well with the visual inspection.