A Method Of Low Contrast Gray Image Enhancement And Its Performance Evaluation Based On PCNNs

This paper presents the theory research of image processing though Pulse-Coupled Neural Networks (PCNN). At the same time, we research the application of PCNN in the image enhancement. This subject which is funded by the fund of nation natural science belongs to the basic theory research. PCNN, which is a new model presented recently and belonged to the third generation neural networks. This paper introduces a new method of low contrast gray image enhancement based on PCNN after studying the operation elements and action mechanism of PCNN. The algorithm achieves the whole enhancement by using the feature of PCNN and the enhanced algorithm of Histogram Equalization. On the other hand, the characteristic of side-suppressing and capture mechanism in PCNN is used to achieve the local contrast enhancement of image. To evaluate our method we use a mess of others methods and some methods of subjective evaluation end objective evaluation. Compared with conventional method, our method has a great effectiveness not only in the overall contrast enhancement but also in the local contrast enhancement of the examined images.