Digital Medical Image Processing And Recognition Technology Research

The1920s witnessed the emerging of the digital image processing technologywhen the first digital photo was successfully transmitted by the submarine cable fromNew York to London. Since then, the digital image processing technology enjoyed rapiddevelopment and has been broadly utilized in many areas including biology, chemistry,medical science, aviation and military science, among which, the communicationengineering and the bio-medical science are the most typical ones. The application ofthis technology in medical science started quite early. Medical images includebio-section microscope images, ultrasonic images, X rays and so on and the imageprocessing technology can be adopted to improve their definition and resolution so as tohelp doctors make more accurate diagnoses in a convenient way. The digital imageprocessing technology, mainly based on computers and other hardware equipment, hasvarious advantages. It is not only highly accurate in processing and rich in content, butalso can deal with complicated nonlinear processing with high flexibility.Beginning with a brief introduction to the development of the digital imageprocessing technology and its current researches, this paper conducts detailed andin-depth studies on different branches of the technology, which include the imageenhancement technology, the image segmentation technology, the image recognitiontechnology. This paper aims to realize the recognition of human’s soft tissue in knee bycoding a recognition software based on the digital image processing technology. Themajor theory is to establish an optimal objective energy function based on the featuresof the digital images in the Gaussian Derivative (image enhancement, edge lifting),defining the edge points by the conjugate-gradient method and then connecting all thepoints together to create an outline of the structure or the organ, so that the imagesegmentation will be achieved. By establishing a Gaussian mathematical model at theearly stage, the paper plans to code a recognition software for the purpose of tissuerecognition.