Research On Tissue Culture Seedling Growth Monitoring Based On Computer Vision

People can get out from the monotonous, repetitive and high-strength work by using computer vision. To monitor by computer vision technology which are improving the data's accuracy and continuity would help people overcome human vision's easy fatigue, big error and other large biological limitations.This paper explored the technology of using digital cameras to get images, with computer vision technology to reconstruct images of tissue culture seeding, obtain the height and the crown width.The research focuses on the image acquisition, digital camera self-calibration, stereo matching and 3D reconstruction. Microsoft introduced the use of digital video software development kit VFW replace traditional analog video cameras to capture images. This paper introduce a new equation based on Kruppa equation and Simulated Annealing Algorithm to calibrate the digital camera, which access to a better robustness. We Structure a new two-step algorithm which choose the walsh Coefficient as matching primitive, first of all, we transform the image gray features into walsh coefficient characteristics in the purpose of preventing noise interference effectively and reducing the matching error ,but it is not to avoid existence of mistake matching. In this paper, we combile the Lmeds and the eight point algorithm to optimize the initial matching, which contribute to eliminate the error matching. As a result ,we increase the accuracy rate of matching. Application of digital camera intrinsic parameters and the motion parameters in the 3D reconstruction to calculate the tissue culture seeding growth parameters.This paper is studied in above algorithms and give a program to verify the feasibility and the accuracy of algorithms which laid the foundation for the next research.