Research On The Image Processing Of Range-gated Dynamic Laser Imaging

Laser slice imaging based on range-gated can obtain both intensity image and rangeimage. Range-gate technique can effectively suppress laser backscattering noise, andincrease the SNR (Signal Noise Ratio) of image. It is one of the few focal plane imagingtechniques that can apply to engineering more easily. Dynamic imaging is an inevitabletrend in the development of laser imaging. In this thesis we have carried out works asbellow on image progressing for range-gated slice imaging system.Firstly, we have researched the image noise. We researched the images’ noise bystatistical analyzing of images without target, According to the characteristics of imagingsystem. We combined several slice images to one composite image, and calculated meanvalue of the noise. Then we obtained the noise’s mean value of single slice image andcomposite image by liner fitting. Results have shown low Gaussian noise of the image.Secondly, we have processed intensity image and range image effectively.Compositing intensity image by mean fusion can suppress low noise. We denoisedintensity image by threshold segmentation, and then segmented range image usingdenoised intensity image as mask, realizing range image denoising. Realized intensityimage and range image composition and denoising.Thirdly, we give an motion estimation algorithm for range-gated imaging, torenovate current image in condition of dynamic imaging. We used gray projectionalgorithm to do motion estimation, and made the current frame and reference framematching through the movement compensation. Then we renovated current image bycomposing the two images. Processing results have shown that this algorithm is notsensitive to the detail of image. But in the condition that images had distortion, thisalgorithm couldn’t realize image renovation.Finally, we researched the distorted image matching and renovation. We did motionestimation and compensation for every slice image, through gray projection algorithm.Effects of image renovation showed that this algorithm realized distorted image matchingand renovation. And if the reference frame had no distortion, the distortion of currentframe was eliminated.