Design And Research Implementation Of Autofocus Algorithm For Leukocyte Microscopic Imaging

Blood smear is an important way to diagnose various blood related diseases.The information of Pathology,such as the number,shape and proportion of leucocyte in blood smear can help doctors to make quick and effective diagnosis.Therefore,the accurate and rapid classification and counting of leucocyte is a direct technical problem facing the medical community and the first problem to improve the efficiency of the hospital.At present,hospitals generally test the types and quantity of leucocyte in the blood by two methods of hematology instrument and microscope.But hematology instrument can not accurately identify the morphology and pathological changes of the cells.Once abnormal patients are found,we must use microscope to make further diagnosis.Although the microscope method can accurately sort,timely detection of morphology and pathological changes,but the need to rely on professional repeated operation microscope visual observation,not suitable for screening large numbers of people,especially the process of focusing imaging microscope,even professionals,still need to spend a lot of time.In view of the above disadvantages of clinical application,we designed and research implementation of autofocus algorithm for leukocyte microscopic imaging.The content of the research is divided into the following three major parts:1)The design of image focusing algorithm.The image focusing algorithm is the core of the research.In this paper,first of all,the evaluation criteria of the focusing function of the 12 existing focusing algorithm models are evaluated,and the focus function curve area is resolving power small,the stability is poor and the larger peak value range.A threshold variance algorithm is proposed for the above problems,and the statistical evaluation is obtained.Standard accuracy error 0.03,peak value range 2um and other excellent curve characteristics.After focusing on the robust analysis of the focused light source,the ratio of the objective lens and the sample of the smear,the gradient threshold variance algorithm is improved and optimized,and the stability,universality and anti-interference ability of the algorithm are further improved.Finally,the focus closed-loop feedback system integrating PID control algorithm and image focusing algorithm is designed to realize the fast,accurate,stable and self focusing motion control system.2)The focusing algorithm of complexity analysis.Due to the higher time complexity of the algorithm and the low efficiency of program execution,with the optimization of the time complexity n scale and the program circulatory structure without affecting the image focusing effect,the experimental results show that the average calculation time of the single original map is shortened to 7ms by 1.42s;the n scale of the space complexity and the cycle program are optimized.After calling the number and internal local variables in the structure,the storage space 83.62 Mbyte temporarily occupied by the original algorithm is reduced to78.26 Mbyte.3)The software and hardware design of the system.It includes the design of the machine,the hardware control circuit and the upper operating software for automatic focusing microscope imaging platform.Among them,system hardware mainly includes: the core control circuit design,computer operating system,lower machine drive system;software includes: PC interface,camera application interface,image processing and so on.Finally,the system realizes self focusing accuracy of 0.2 um,and self scanning for 150 leucocyte images,are systematically tested and evaluated.After repeated experiments,automatic image focusing takes about 2 seconds,and automatic image scanning takes about 4 seconds.The whole operation cycle of the system takes about 7 seconds on average.Finally,the design and research implementation of the autofocus algorithm for leukocyte microscopic imaging are completed from the design of image focusing algorithm,the focusing algorithm of complexity analysis and the software and hardware design of the system.A leukocyte automatic screening and screening system which can improve work efficiency,ensure detection quality and high practical value has been realized.