The Image Retrieve Based On FPGA

With the rapid development of Internet and the rapid growth of image information, extracting valuable knowledge quickly from massive collections of digital images has become the people's urgent needs. Since the 20th century, 90's, the content-based image retrieval (Content-based image retrieve, CBIR) technology came into being, which becomes hot spot in the field of intelligent information processing research.In the filed of robot vision, fingerprint identification, medical and remote sensing image analysis and processing, content-based image retrieval technology has a wide range of applications in real life. The practical application of image retrieval have become increasingly demanding of real-time and embedded environment. So, how to achieve image retrieval technology in the embedded environment has become the new direction of development.FPGA (Field Programmable Gate Array) provide hardware guarantee for the rapid development of embedded image retrieval system.FPGA is a high-density programmable logic device, which has a wide range of applications in the digital logic systems. It can satisfy the situation which need to be taken into account the speed and flexibility of applications.The article studys the FPGA technology, EDA technology, hardware description language and modular design methods, and then expounds the current popular content-based image retrieval-related technology.In order to solve the problem that a single image feature vector for image database query can not solve the contradiction of the accuracy and efficiency, the article designed a comprehensive multi-dimensional feature vector hierarchical search algorithm. In order to solve the problem that the accuracy of traditional color histogram retrieval algorithm is low and the efficiency of fuzzy color histogram algorithm is low, the article designed a histogram Gaussian filter retrieval algorithm.Upon the completion of the image retrieval algorithm, in order to farther improve the efficiency of image retrieval, and make image retrieval to be used in broader embedded environment, the article studys the hardware design of the Gaussian filtering of the histogram retrieval algorithms. Through hardware platform design and system software design, the FPGA-based image retrieval system's hardware implementation is completed.The study of Color-based image retrieval algorithm and hardware implementation of image retrieval has important practical significance, it also has a good reference effect for large-scale implementation of embedded image retrieval.