Research Of Recognition Technology On Blood Cell In Urine

Urine occult blood cells is a characteristic whether the health is well，as well as it is alsoone of the important items for the clinical medicine detection routine.With the development ofintelligent detection in urine cell technology and the automation of medical testing, the cellrecognition technology in urine testing has gradually become possible, based on it, applicationof technology has increasingly become a focus of research. Maturity and development of thistechnology will facilitate research on the level of clinical laboratory medicine, which has avery important theoretical significance and practical value.Through a comprehensive analysis of advanced technology of the current stage at homeand abroad, according to the latest findings,this paper raises the urine cell based on supportvector machine classification method automatic identification technology and makes acomplete classification system.In the process technology,it fuses and utilizes biomedicaltechnology, the advantages of computer image processing and computer vision technology,by analyzing the characteristics of the image itself, pretreatment urine occult blood cell imageand determine the characteristics of the cells.In the extraction, classification and identificationof cell characteristics, the performance of system algorithm has been improved,by usingsupport vector machine parameter optimization and optimal exploration.It compares the useof urine cell identification and effect of classification results of support vector machines in theRGB and HSI two different color space, makes the analysis of basic algorithm and its textureparameters better, improves the accuracy of urine cells image recognition.It describes amulti-class support vector machine in three segments method, compares and analyzes theeffect of urine cell recognition and identification rate of the different methods.Experimentshows that it has a good effect which is the LBP algorithm improved proposed approach andmulti-class support vector machine effect on urine cells recognize classification.