Research On The Image Fusion Technology Of The Stereo Microscope

With the development of the computer and image processing technology，the traditionaloptical microscope can not meet the needs of researching. The digital stereo microscope arises atthe historic moment. The stereo microscope system has been widely used in micro manipulate,micro measurement and so on, and can also be used to solve the problem in the military andaerospace. In the imaging processing of the digital stereo microscope, the stereo images can not becaptured clearly in the two channels, because the depth field of stereo microscope is limited. Inorder to obtain a clearly image for comfortable and convenient, the intelligent processing should beadded. The fusion method can effectively solve the problem, which can make up the defect. In thispaper, the fusion algorithms of the stereo micro-images were mainly researched. Two new fusionalgorithms are proposed to overcome the shortcomings of the current image fusion algorithms,which are not suit for the stereo micro-images. Then, two image post-processing algorithms areproposed to deal with pseudo boundary.Firstly, a stereo micro-images fusion algorithm based on the region growing method has beenproposed combing with the region growing model. This algorithm gets sub-region feature valuethrough stereo micro-images definition evaluation functions. Then we can obtain monocular stereofusion micro-image through the fusion rule. The results, which are compared to some fusionalgorithms, show that the algorithm based-region growing is effective and superior to otheralgorithms.Secondly, the second fusion algorithm is the maximum expectation technology for the stereomicro-images, which is based on expectation maximum technology. This proposed algorithm isadaptive without selecting some parameters and has superior quality in holding features of sourceimages. Here, this algorithm is applied to the monocular stereo micro-images fusion. Theexperiment results show that the algorithm achieves good performance. At the same time, thisalgorithm is applied to the binocular stereo micro-images fusion and the experimental results showthat the proposed algorithm is successful for stereo microscope fusion images.Finally, the pseudo boundary will appear in the fusion of stereo micro-images, which isobtained through the region fusion method. In order to solve the problem, two image post-processing algorithms are proposed: the color weight filter interpolation algorithm and the colorgradual change filter interpolation algorithm based on CIELAB color space model. The experimental results show that the proposed two algorithms can solve the pseudo boundary andachieve good effect.