Wavelet Theory-based Nonlinear Image Processing

The approach of traditional image processing technology is differ from human vision (HVS), so the quality of image is not very well. The similarity of wavelet analysis and HVS is a unique advantage in image processing.Wavelet analysis has a very wide range of applications in signal processing as an important tool for Time-frequency analysis. This paper focuses on the wavelet transform theory in the non-linear image processing.The fundamental theories of wavelet analysis are discussed in detail. Continuous wavelet transform, discrete wavelet transform and multi-resolution Analysis transform, the fast algorithm of discrete dyadic wavelet transform-Mallat algorithm, lifting wavelet transform theory, integer wavelet transform and multi-band wavelet transform theory are introduced, and the role in image processing of these theories is analyzed.The cause of jagged edges and fuzzy effect and the performance of the traditional multi-focus image fusion algorithm is analyzed. On this basis, image nonlinear processing based on wavelet transform is introduced combining principles of HVS.The algorithm is detailed analysis and designed. Experimental results show the effectiveness of the algorithm.Wavelet transform is able to maintain the details of the image-rich information. Binding the experiments, this algorithm presented in this paper got not only a good objective evaluation of the data but also a good image subjective quality. This finally validated the effectiveness of the algorithm. This finally validated the effectiveness of the algorithm.At last, this paper designed a simple and effective algorithm which is more conducive to FPGA implementation through the analysis of the FPGA design principle and data pipeline of algorithm. This algorithm increases the hardware utilization.