Research On Forest District Extraction Algorithm Applied To High Resolution Remote Sensing Image

Forest resource investigation and monitoring is the main content of studying vegetation coverage comdition of the earth’s surface and global ecological change currently. And the investigation and statistics of the forest area, which is considered to be the basic content of forest resource investigation and monitoring, turns into just work of the paper. To study and solve the problem of estimating forest area is also to do the problem of forest classification in high spatial resolution remote sensing image. To classify the high resolution remote sensing image and further to extract the forest part, are the concentration and main content of the paper. The current classification in high resolution of remote sensing image is to process the pixel data mainly by the computer as the tools, so as to achieve the purpose of classifying geomorphic characteristics. Clustering analysis technology is the main research content for its unique classification advantage. With the continuous development of new technology, clustering analysis methods emerged in endlessly from the traditional K-Means algorithm to the Fuzzy C-Means algorithm. Although these methods improved the remote sensing image segmentation to some extent and promoted the application of remote sensing technology, different algorithms had their advantages and disadvantages. According to the fully study and analysis of the domestic and abroad research and development of classification methods in high resolution remote sensing image, the paper proposed an adaptive weighted fuzzy c-means segmentation algorithm, and made use of the fractal dimension characteristic to extract the forest area after segmentation. The experimental results show that this method has got good proformance in image segmentation and forest region extraction. It has not only increased the segmentation accuracy of remote sensing data, but basically realized the automation of data processing.