Research On Target Recognition And Tracking Algorithm For Minimally Invasive Nuclear Transfer

As the core technology of animal cloning,nuclear transfer has been widely used in pharmaceutical,medical,endangered animal protection and other fields.At present,there are many disadvantages such as low efficiency,low precision and low success rate existing in manual operation.In order to improve the accuracy and success rate of nuclear transfer operation,this paper aims to explore an algorithm for target recognition and tracking during nuclear transfer operation.Firstly,by reading a lot of literature,a deep understanding of micromanipulation system,nuclear transfer operation and visual detection and recognition technology has been formed.And then,the advanced technical methods used in nuclear transfer and the disadvantages such as low precision and large cellular damage in the nuclear transplantation are analyzed this paper.Therefore,the problems needed to be solved urgently in the nuclear transfer microsurgery are clarifiedSecondly,based on the analysis of the operation process of nuclear transfer,the operating system of nuclear transfer using piezoelectric vibration to realize membrane rupture is introduced.According to the operation requirements of nuclear transfer,the function of micro visual system is analyzed.Based on the function analysis,the vision system which composed of the polarized light imaging system and the inverted microscope is established.Thirdly,the method of microscopic target recognition in nuclear transfer is studied.To assist the oocytes' screening,an automated zona pellucida thickness measurement method is proposed this paper.According to the requirement of micro target localization in nuclear transplantation,the localization algorithms of oocyte,nuclear material and somatic cell are studied.Experiments are carried out on several images.According to the result,the oocyte localization success rate is 100%and the success rate of nuclear localization is up to 96%.Then,the real-time tracking method of the injection needle is studied.Based on the analysis of the template matching method,a real-time tracking method of the injection needle based on OR B algorithm is proposed.And then,the algorithm is tested by using the common used injection needle(flat needle and sharp needle)from two aspects:detection speed and detection precision.The experiment result shows that the processing speed of the algorithm is 22 frames per second,and the detection error is about lum.Finally,an experiment is taken to study the proposed target recognition and tracking algorithm during nuclear transfer operation.The thickness of the zona pellucida is measured by two experiments:different positions of the same oocyte and different oocytes.And the experiment result of the zona pellucida is similar to the empirical value.Moreover,the zona pellucida thickness measurement algorithm,target localization algorithm and micro needle tracking algorithm are researched and applied in the oocyte's screening,cell immobilization and pose adjustment,enucleation,nuclear injection and so on.