Research And Implement Of An Image Processing Arithmetic Evaluating Flatform

With the development of computer, multimedia and data traffic technology, digital image processing is used in almost every industry. And with the development of information technology, it requests the data transfer speed higher and higher, the data volume larger and larger. The real-time image processing has made a broad attention. However, the processing flatform before can not satisfy the transfer speed nowadays.Investigated the situation of real-time image processing, a high-speed digital image processing and evaluating flatform is built in this thesis to adapt to higher and higher requests. First, it designs the whole frame for the system, describs the configuration, technical solution, equipment requirement and so on. To the system of data collection and transmission, the paper analysis of the structure and characters of the infrared imaging devices and the relevant bus interface, and then, identifies the design of internal and external interface board. Besides, under the requirements of real-time image processing, store-and-forward mechanism is designed for the USB interface transmission to prevent the transmission of dithering and the indefinite delay.In the program processing, this thesis focus on the software of system. The software support includes the essential software drivers to support hardware and the display software on PC. The display software is designed base on software engineering of RUP model, including the receiving, broadcasting, storage of the images and data that original and processed by hardware system. Besides, according to the particulality and complexity of imageing envirionment in this project, the paper is focused on the restoration method of motorial fuzzy image and makes experiment for each. And then, compares the restoration results, analysis the limitations of the basic restoration model, puts forward the algorithm of improving method and direction.