Synthetic Vision Enhancement Technology Research Based On Multi-source Sensors

Due to the complexity of natural environment, the low-visibility weather conditions such as fog, rain, snow, and other poor visibility. Therefore, the technology of vision synthesis for multi-source images enhancement has great significance. The enhancement technology uses sensors and advanced technology, using the sequential method to fuse images and obtain the maximum amount of information to perceive the environment, for example, the driver's vision is enhanced in snow or dark light conditions of the environment for visual effects, it can auxiliary drivers to drive safely.Image enhancement means improving the image's quality, its purpose is to improve the visual image by enhancement algorithms. Traditional visual enhancement is often targeted, so that good results on certain types of images may not be suitable for another type images, with image enhancement technology deepening, the new image enhancement methods are emerging, information fusion has become the primary means for image enhancement. This thesis's research work is mainly carried out around the vision synthesis enhancement, the contents and results are as follows:1.The thesis studies the traditional image enhancement methods, analysis visual enhancement algorithm's defects and shortcomings, propose synthetic multi-source sensor visual enhancement technology compensate for its shortcomings.2. Trandition pulse coupled neural network is complex, improve its model to improve network speed, and use the local energy rate into neural network to obtain better fusion results.3. Analyses the multi-source sensor image's visual characteristics. Compare pixel-level fusion algorithms and propose a new algorithm based on the weighted feature energy algorithm for multi-source fusion, using this fusion algorithm to achieve the purpose of enhancement.4. Under above theories, methods and techniques comprehensively for multiple comparisons test and give analysis of the fusion quality evaluation that shows this method can obtain better vision synthesis enhancement effects.