Implementation Of Image Enhancement Algorithm

Digital image processing has developed quickly in recent years, it is a new kind of research field, more and more scholars pay attention to it. This paper aims at digital image enhancement as the object of study. It is an important branch of digital image processing. During the generation, transmission, compression, storage, transformation and other a series of process, image will be affected by various factors, such as: variations in the operating environment’s illumination can cause lighting changes; when operating the equipment, shock inevitably occurs, causing displacements; it is common to capture image with low contrast or inappropriate position, etc. In practical application, these factors can cause image quality degradation, affect the overall visual effect of the image. Therefore, in order to solve these problems, it is need to enhance the image, improve the quality of the image. Image enhancement is an important part of image processing, aim at enhance the image with the purpose, for example, it is not only enhance the contrast of entire image or the local details but also inhibit the unnecessary details of the image, make fuzzy image turn into clear image, improve the quality of image, so that in the subsequent image analyze, an operator or a machine vision system can be better to analyze and understanding the image.Firstly, this paper introduce the commonly methods of image enhancement, including the principle and method of the image enhancement algorithm in the spatial domain, such as gray level transformations (image of reversal, contrast stretching, log transformation and power-law transformation), histogram processing (histogram equalization and histogram specification) and spatial filters (median filter, averaging filter, Laplace filter, Gauss filter and Sobel filter); the principle and method of the image enhancement algorithm in the frequency domain, such as Ideal lowpass (highpass) filter, Butterworth lowpass (highpass) filter, Gauss lowpass (highpass) filter, and homomorphic filter; the method of combining image enhancement.Secondly, based on the methods of image enhancement in this paper, author make use of the user interface provided by MATLAB software platform established three modules, including image enhancement in the spatial domain, image enhancement in the frequency domain and image operation. Then introduce the design and main frame of the software and implement methods of separate module. The software fully use of controls and internal functions of image enhancement provided by MATLAB GUI module, achieve the software development of digital image enhancement, and realize the algorithm of image enhancement. Finally, with the software of image enhancement, author has realized the process and analyzed of the typical images in different field. In this paper, author mainly realized the methods of image enhancement in the spatial domain, including gray level transformations, histogram processing and spatial filters; the methods of image enhancement in the frequency domain, including lowpass filters, highpass filters (add offset and coefficient to the filter, realize the high-frequency emphasis filter) and homomorphic filter; the combining image enhancement method, combine the image arithmetic module (a linear combination of the two images, subtraction, absolute difference, multiplication, and division) with aforementioned methods of image enhancement, then achieve the combining image enhancement.The software programming of implement on image enhancement algorithm has been used in digital image processing, it turned out that the design of the interface is friendly, the feasibility and reliability of the software has been verified, and it also has better visualization operation process.