A Study On The Technology Of Information Processing In Optoelectronic Imaging System For Biology

The optoelectronic imaging system for biologic specimen can obtain high qualitative biologic image and realizes the aided analysis of the imaging object by taking advantage of the compute information technology and optoelectronic imaging. Information processing techniques are very important in optoelectronic imaging process such as imaging control, image construction, image enhancement, and image analysis.The optoelectronic imaging system is composed of optical imaging, photo detector and image construction. It works through a set of system software. The theory analysis, algorithms research, numerical experiment and engineering application about adaptive image enhancement, blind image restoration and microscopic image analysis is emphatically studied in the dissertation.Biomedical image is of non-stationary spatial statistics. The enhancement algorithm based on stationary spatial statistics can't track the variations of the image statistical characteristic. However, tracking this variations is just advantage of two-dimension adaptive system. A system based on TDLMS can realize biomedical image adaptive enhancement and track the variations of statistics property at the same time. It not only enhances image effectively, but also retains the image's details. The numerical experiments results are given and verify the above conclusions.Blind image restoration may recovery degraded image and also estimate PSF. At present, developed algorithms for blind image restoration are constrained to some extent in applications, which depends on consistency of application conditions with algorithm assumptions. In terms of optoelectronic imaging for biologic specimen, a novel algorithm, called frequency domain conjugation gradient algorithm (FCG), is presented in the dissertation. We establish a new cost function in frequency domain based on imaging information. The conjugate gradient algorithm is used in the minimization of the cost function. The physical optimization result is obtained under the constraints imposed by imaging information. Numerical experiments are made to simulate the degradation arisen from diffraction and defocus in optoelectronic imaging system for biology. The deblurred image and the PSF are well simultaneously estimated by using this algorithm. Finally, the dissertation discusses the technology of microscopic image biologicalfeature analysis and recognition in pharmacological experiment research. A set of algorithms and program is developed for quantitative analysis for of morphology in immunopharmacology, which is actually used now.