Research On The Key Technologies Of Solid State Nuclear Track Image Measurement

With the enhancement of domestic environmental monitoring and protection, high precision measuring method of nuclear track image are more widely used by the use of modern image processing technology and optical technology.The principle, development and application prospect of solid state nuclear track detector (SSNTD) are introduced systematically in this dissertation. By analyzing the previous research results of nuclear track image measurement technology and the influence factors on the track image measurement accuracy, a set of new methods of track image filtering, enhancement, edge segmentation and recognition are established. The blurring,deformation and overlapping of track image have been solved in the track image measurement. And the accuracy and efficiency of the track image measurement are improved. The results are verified by related experiments. The main work is as follows:1. The problem of over segmentation and less segmentation to the overlapping nuclear track are caused in using the usual segmentation algorithm. After analyzed the causes of noise and uneven brightness of track image, the improved algorithm of denoising,enhancement and segmentation are determined to solve the problem. Firstly, the local histogram equalization method is used to deal with uneven brightness. Secondly,morphological filtering are used to smooth the image and fill the hole in track center.Finally, the nuclear track image are segmented by using the OTSU or the improved watershed segmentation algorithm.2. Aimed at the ellipse target recognition question in the track image recognition,Hough transform method is improved by using elliptic center geometric features and completed the calculation in the two dimensional space. The measurement of track image is completed in pixel level.3. With the development of modern image technology, image recognition method of nuclear track are improved by the pixel level to sub-pixel level. A variety of image sub-pixel extraction method are analyzed in this dissertation. And combined random and the root mean square(RMS) concepts, the ellipse parameters in sub-pixel level are obtained by adopted least square ellipse fitting method.4. The problem of track overlapping seriously affected the measurement accuracy. A new method of separating overlapping nuclear track is proposed based on image matching and twice erosion operation, according to the features of track image. The basic idea is that non overlapping track is used to identify and isolate the overlapping track. The shape matching and twice erosion operation are used to separate the overlapping nuclear track.Firstly, non overlapping nuclear track are detected and identified by the area and circularity in this method. Secondly, the non overlapping nuclear tracks are used as template structure elements. And the twice erosion operation are used to separate the overlapping nuclear track. Finally, the classification images are formed. The experimental results show that the track image are segmented effectively and accurately by using the shape matching and regulated morphology.5. After analyzed the various factors of affecting the autofocus accuracy, to eliminate the influence of the backhaul error, an automatic focusing control strategy is proposed based on function steepness and variable step. By reduced the number of reverse movement,it reduces the backhaul error and achieves the coarse and fine auto focusing.6. The optical system inclination caused the deformation and quality degradation of microscopic track images. A realtime method of measuring the image inclination Is presented on an improved laser triangulation. The basic idea is that the traditional laser triangulation is improved by using multiple beam parallel light. An optical system is constructed based on multiple spot and laser triangulation. And the relation between the rotation angle of the mirror and the displacement of CCD spot is deduced. It could detect the deviation of microscopic image and meet the requirements of image measurement accuracy in track image measurement. It lays a good foundation for the following microscopic image processing and stitching.