The Research And Application On Rice Detection System Based On Image Processing

There is exigent demand that machining of grains is asked not to stop while those grains being measured and that control should be feed back the machining equipment in real time in some enterprises. This thesis emphasizes only on resolving application envelopment of measurement system on line to machining of industrial grains. The thesis brings forward using computer image measurement technology to acquire and process and analyses image of industrial grains in real time. Traditional handcrafted manner will be replaced by this system. The system will make operator's working intensity descend and enhance efficiency of measurement and analysis, moreover, output corresponding analysis data and curves and images and diagram according to user's need.The thesis tells of theory and technology about application development, and introduces system structure and approach of achievement, and presents design tentative plan of system hardware, including the selection of kinds of equipments, especially those algorithms of this system are developed and applied. Image-processing is the most key part of Rice-detection system, moreover, is also important content of the thesis. According to some standard parameters of this system, rice machining appeals and inherent characters of rice, many better application value algorithms of image process are presented with ideal images and factual sampling ones. Some algorithm, including optimizing threshold for the making image two value and volume product for image enhancement, new adaptive median filter algorithm, a algorithm for image binary-conversion based on mathematical morphology, the application of morphology-watershed in segmentation of overlapping grains, a rice binary image thinning algorithm based on critical pixel reservation and seed-fill algorithm for two value image processing and grains measurement, are key in system.Chapter one introduces background of development and status of application and development of computer image measurement technology. Chapter two tells of technology and theory, which will be used in system design. Chapter three emphasizes on analyzing and application of convolution algorithm. An filter algorithm that an adaptive median filter for rice-detection system is presented, combining the merits of IMFLED algorithm and EM algorithm. Chapter four comes out the application and development of computer image segmentation, including image binarization algorithm and overlapping grains segmentation algorithm. Chapter five presents binary image thinning algorithm. Based on thinning principia, thinning algorithm based on critical pixel reservation and thinning algorithm based on morphological mathematics are presented. These problems of framework asymmetry, edge gap phenomena and edge-width are solved with these two thinning operation. Chapter six presents development software and hardware of image processing. Chapter seven sums up the application develop and put forward advice for system betterment.