基于fpga的高速图像预处理的算法研究与实现仪器科学与技术专业论文.docxVIP

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 - - II - Abstract Image processing is widely applied in many fields, such as military affairs, remote sensing, medicine and industry. The fast growing quantities of the data of the image processing and the requirement of the real-time operation ask the designers to change the algorithms of the image processing and use some more effective devices. The preprocessing of image researched in this paper is about the low-grade of image processing, which often deals with a lot of data and with simple algorithms. And the result of the preprocessing of the image can be used by the following equipments. FPGA is suitable for the preprocessing of image. In the last twenty years, FPGA has developed very fast. Now, with its high integration degree、low power consumption and high flexibility, FPGA is very suitable for users to design and change some function of a system. Median filtering 、 maximum filtering 、 minimum filtering 、 edge detection、histogram equalization are very useful image preprocessing methods. This paper researches the algorithms of these preprocessing methods and realizes them on FPGA. In the design of median filter, an improved algorithms of median filtering is given, which can help to improve the efficiency of the processing. Maximum filtering and minimum filtering are two useful ordinal filtering methods. In this paper, the algorithms of maximum filter and minimum filter are improved, which can reduce the usage of FPGA resources and reduce the deal time of the operation. Prewitt detection、Sobel detection and Isotropic Sobel are three very useful edge detection methods. In this paper, the algorithms of there edge detection methods are given. Histogram equalization is an important image processing method. This paper researches the histogram equalization algorithm based on FPGA. The algorithms of the image preprocessing methods mentioned above are emulated.The results of the emulations accord with the design. Keywords: FPGA