Research On Reservoir Target Multiband Color Imaging Technique

Numerous water vessel transportation accidents have happened in the reservoir region in recent years. In fact, the aquatic traffic safety emergency response still faces a stern tendency as a whole. At present, the priori task of solving aquatic traffic safety lies to get a general understanding of vessel movement in the reservoir region, discover the abnormal channel, carry emergency rescue effectively and timely. Hence, the effective imaging of targeted water area is the key. However, the existed Single-band imaging technology has undesired imaging effects, such as Infrared thermal imaging technology, Low level imaging technology. When facing the safety accidents, it could not carry an effective emergency rescue and could not satisfy the imaging requirements under different weather conditions as well, likes fog and rain. Therefore, it is an emergent task to research an ideal imaging technology which suits for water monitoring in the reservoir region.Multiband Imaging Technology is defined that use the sensors of different bands to form the image for the same targeted scenes. Then it gets the high level event detection technology via image fusion scheme. This technology overcomes several weakness, likes the low information of Single-band imaging technology and low level event detection. In addition, it can obtain some useful information of several bands. Plus, it has good target discrimination ability, high fault tolerance and good detect ability. Thus, Multiband Imaging Technology is more suitable for water imaging system in the reservoir region.This dissertation is based on the existed image acquisition and processing technology. According to the requirements of specific project, a Multiband Color imaging technology system in the targeted reservoir region would be designed, which lies on the Multiband Color technology. Infrared thermal imaging technology and Low level imaging technology form sensor array, then it shapes the image of the same targeted water area. DSP Image Acquisition Board not only done the task of image acquisition, but also makes the image data pretreatment, edge detection and partitioned color conversion. So, it will get more perfect pictures of water area. In addition, focus on the weakness of unclear partitioning algorithm and low information, this dissertation is aimed to design a method of partitioned color conversion which strengthens detailed information of pictures. Furthermore, it is also consistent with visual effect, improves image sharpness and more convenient for event detection.Through the primary joint debugging of system's software and hardware, it proves the technology of this system could be achieved. This thesis has done the task of multiband image acquisition treatment and achieved expected effect. Plus, it lays a solid foundation for further site debugging and application.