Studies On Enhancement And Edge Detection Of Remote Sensing Image

With the development of the remote sensing technology, the analysis and processing of remote sensing information has also been considerable development. Different from industrial image and medical digital image, the type of remote sensing image is more diverse and more complex content. Therefore it’s necessary to do some preparation before analysis, such as using some algorithm to achieve the enhancement of remote sensing images, improve image’s contrast, highlight the details of the image, and the information of the image’s edge is also very important information.According to the features of remote sensing images, using the wavelet transform, nonlinear filtering, histogram equalization, Otsu threshold, artificial fish swarm algorithm and genetic algorithm for remote sensing image enhancement and edge detection. The main research contents are as follows:1. Study on the enhance arithmetic of the orthogonal wavelet transform and nonlinear filterIn order to facilitate the processing of image enhancement and make out the distribution of the image’s information, by introducing the biorthogonal wavelet transform algorithm that decompose an image information into the low frequency coefficient, the high frequency coefficients in horizontal aspect, the high frequency coefficients in vertical aspect and the high frequency coefficients in diagonal aspect. By using the nonlinear filtering algorithm to transform the low-frequency and the high frequency coefficient after the decomposition, the most of the noise points in image are removed. In the process of image enhancement, not only retains the useful information of the original image, but also restrains the diffusion of the noise points in the image. In this way, the result of enhancement processing is relatively in reason.2. Study on the enhance arithmetic of the histogram algorithm and Otsu threshold algorithmUsing the histogram algorithm instead of histogram equalization to enhance the image, it is because the histogram equalization is aimed at the whole image and the noise is also amplified in the process of enhance. While using the histogram specification, can effectively control the distribution histogram, most of the noise points are excluded, and keep more information of the target. The introduction of Otsu threshold makes the remote sensing image enhancement effect is more significant, it is because the Otsu method is adaptive and have the strong search capability.3. Research on the remote sensing image edge detection algorithm based on artificial fish swarm and seed growthThe artificial fish swarm algorithm is a typical intelligent bionic algorithm. By using the rule that activity or control the biology’s behavior to realize this algorithm. The artificial fish swarm algorithm is simple, parallelism, fast and global. In the process of using the seed growing algorithm, as the ability of artificial fish swarm to search for the optimal solution, in this way, we can find the better edge information. So the results of using this method obtain the better edge detection.4. Study on the edge detection that based on mathematical morphology and genetic algorithmGenetic algorithm is also a kind of intelligent bionic algorithm, which is based on the theories of biological evolution. It is mainly related to the biological heredity, variation characteristics and so on. As the genetic algorithm also contains a crossover operator, therefore, genetic algorithm has the collateral characteristic, and also has the good optimization ability. At the same time, by using the typical method corrosion and expansion of mathematical morphology algorithm to get the edge detection results, this method obtained a better results.