Yinguo Huang

Application of contourlet transform in infrared image denoising

Wavelet threshold denoising is widely used in the denoising of the infrared image for its simplicity and effectiveness in application. However, there has been a growing awareness to the observation that wavelets may not be the best choice for describing infrared images. This observation is due to the fact that wavelets are blind to the smoothness along the edges commonly found in images. A denoising method of infrared image based on Contourlet transform is presented in this paper. In selecting the hard threshold function to process the coefficients in the Contourlet domain, we could thereby obtain the denoised infrared image of superior quality via inverse transforming. The result of the experiment indicates that compared with the traditional algorithms of the wavelet, this method can preserve the detail and the texture of the infrared image more effectively, and has better image effect and the SNR value.

Infrared and visible image fusion algorithm based on Contourlet transform and PCNN

A new image fusion method based on Contourlet transform and an improved pulse coupled neural network (PCNN) is introduced in this paper. The input infrared and visible images are processed with Contourlet decomposition which has multi-scale and multi-directional characteristics. The PCNN algorithm deriving from the neurophysiology is optimized in order to be compatible with the image fusion strategy. Owning to the global coupling and pulse synchronization characteristic of PCNN, this new fusion strategy utilizes the global features of source images and has several advantages in comparison with the traditional methods based on the single pixel or regional features. Multiple criteria and statistical indicators regarding different aspects of image quality are presented for objective and quantitative evaluation of the fused images to understand the performance of image fusion algorithms. Experimental result shows that the new method can improve the quality of image fusion and can achieve an ideal fusing effect. The method would find its application in the aspects of optical imaging, target detection and safety monitoring, etc.

Research on laser interference alignment based on interference fringe

A new method of straightness measurement is developed in this paper. It makes interference beam be the reference axis in order to measure the straightness of rail or other objects under test. This method can reduce the measurement error brought by the thermal distortion of laser source and environmental factors in a certain extent, and the accuracy of straightness measurement is advanced.

Study on image matching of 3D measurement system on dynamic object in long distance

The binocular stereo vision measurement system with two CCDs, whose optic axes are parallel, is presented in this paper to measure the 3D profile of a dynamic object in long distance. Two key technologies in the binocular stereo vision measurement system are dissertated. The first technology is to extract the features of two images which are separately gathered by two CCDs. The second technology is to process the feature pairs to build a corresponding relationship of different feature points of the same space point and get the disparity image. Experiment results of stereo matching and 3D profile of measured object are given. These results show that the mathematic model is simple relatively, stereo matching is easy comparative, and the measurement is rapid. And the presented binocular stereo vision measurement system can be used to measure the attitude and 3D profile of flying missile, and the measurement system which requires some hardware, can be realized easily, operated stably and is of high measurement precision.