The Research And Application Of Holographic Recognition Method For The Micro-image

In recent years, the rapid development of computing technology, for computer imageprocessing technology development to provides a broad platform. The computer imageprocessing technology in all walks of life is playing a more and more important role. Theresearch of microscopic image recognition method is the main subject in image processing;many domestic and foreign researchers propose different identification method for themicroscopic image recognition. This microscopic image with particles as the research object,the key research is the particle image holographic recognition.In the microscopic image processing, the holographic recognition we defined as: themethod contains the geometrical characteristics of the particles and color informationrecognition. This topic bases on color model theory, the particle image of high precision, highspeed and holographic recognition are the goals of the topic, we create holographicidentification method and provide a fast algorithm. We use the transformation of the colorspace, in order to finish the identify of particle curved surface piece on the surface, the imagerecognition method bases on2d space and leads to based on3d space, from edge detectionand recognition leads to the surface and domain holographic characteristics identification, It isfor the application of image recognition technology to provide new ideas and methods.First of all, according to the goal of the subject requirements we construct themicroscopic image recognition system framework of holographic namely system technologyline. The development of the system we choose Visual Studio2010asp.net VC++for theprogram design language. We analyze the key technology of system development IUV colormodel, we analyses the system of the technical route for feasibility analysis, and completesystem function module design.We must realize the microscopic image of geometric features recognition first to do is torealize the color image edge detection. Active contour model proposed for the traditionalcomputer vision theory and application research has brought new mode of thinking andsolutions, more and more applications and research show the effectiveness of the method. Butactive contour model on the initialization speed, the algorithm convergence speed and algorithm on the stability effect the accuracy and speed, we improved the original activecontour method, so the identification precision and speed methods improve, it can completethe particle image edge detection. The extraction of geometric features of the particle is theultimate goal we particles edge detection. Usually, the geometric features include thecharacterization of particles of particle size parameters and characterization of the parametersof the particle shape; the particles of the geometry information have great significance for thepractical production.Researching the microscopic image of color characteristics identification method, theobject and the texture pattern division on the surface cause more and more people’s attention,in the surface we use the most close to realize method in the real meaning of color imagesurface characteristics identification. First of all we create surface according to the IUV colormodel, the most close to point method is to use the closest point to represent surface, andtherefore it does not need to parametric design, also does not need closed surface, and doesnot even need to know surface direction. This is its remarkable characteristics, we useholographic recognition method to solve a major problem in particle image processing of isthe adhesion of the particles segmentation problem.The application example of system, microscopic image analysis analyzer is the particlesize and distribution of the special apparatus with particles morphology. The instrumentthrough the well-designed optical microscope and a professional camera to obtain particlesmessage; analyzed and processed by using computer image analysis technology, thus we candetermine the size distribution and morphology of particles. According to the requirements ofoutputting analysis report of the static analysis and shape analysis of particles, we develop asystem and introduce in the passage.