| The 21st century is the era of ocean development, which has now become global consensus. In the situation of rapid population expansion, land resources depleted, the deteriorating environment, the development of ocean resources is a far-reaching impact on the 21st century strategic choice.Underwater Image analysis and recognition is one of the key technologies of the ocean research and development.However, due to complex imaging conditions ang poor quality,underwater image is difficult to be used for marine engineering, marine scientific research and military fields directly. It has caused great problems to the actual operation and scientific research . Therefore, the underwater image is needed to make the corresponding enhancement process, then it will be more conducive to high-level image processing , more in line with the requirements of human's vision and machine identification.This paper focuses on the study of image enhancement algorithms and its applications in underwater image. the principal tasks are as follows:First, based on the basic theory of image enhancement , this paper introduces the current several major image enhancement methods detailed.Compared them,this paper analyses the practicality and limitations. Then, this paper expatiates the optical properties and imaging principle of underwater image, and provides the theoretical basis for the research .Second,This paper introduced in the equal power spectrum-homomorphic filtering method to enhance actual underwater image for the first time.Homomorphic image filtering method can eliminate the problem of uneven lighting, enhance the image details. The equal power spectrum filter can effectively restrain low-frequency component, and retain more high frequency components, this is suitable for underwater image. For poor contrast of details and blur, underwater images, which was filtered by the equal power spectrum method, will have symmetric brightness and the details can be enhanced after processing.Then , with the summarization of above content , based on the imaging 's rule and characteristics underwater image, this paper makes a corresponding improvement of the equal power spectrum filtering , which combined the advantages of wavelet decomposition .The improved method chooses an more appropriate system transmission function to retain more high-frequency components and enhance local details effectively. Experimental results show that the improved method this paper presented, can help increase the detail clarity of image ,and give better outline. The enhanced effect is been improved further. |
