Study On The Key Technology In Imaging System Of Inland Vessels Scene Of The Accident

Water transportation in China has made rapid development since the "Eleventh Five-Year". Because of inland shipping business is becoming increasingly busy in recent years, inland ship accidents have repeatedly occurred in China. Inland ship accidents are sudden and always occur in adverse weather conditions or in periods of poor illumination conditions. Poor Visibility makes the work of searching and rescuing in scene of accident difficult to finish quickly and efficiently. Therefore, whether the ship scene of accident can be imaged clearly and accurately significantly influences the success or failure of the work of searching and rescuing in the scene of the accident of inland vessels.Currently, most of imaging system for water target use single-band imaging technique. Single-band imaging technique has its own limitations, unable to meet the imaging requirements of inland vessels scene of the accident, which need to accurate and clear imaging throughout a whole day. Multi-band imaging is imaging on the same target scene at the same time by image sensors of a number of different bands, it combines imaging characteristics of these different bands, through a series of image algorithms, finally output an image contains more information and is more suitable for observation by human eyes. It can overcome the shortcomings of single-band imaging technique which subject to weather and illumination conditions affect, and can imaging accurately in different weather and illumination conditions. Because of its high amount of information and strong ability of anti-interference, multi-band imaging technique is applicable to inland ship accident scene imaging.The paper aimed on the project requirements and the characteristics of the water target imaging, designed and completed a specialized imaging system which based on existing image acquisition processing techniques and multi-band imaging principle for inland vessels scene of the accident.It designed and developed hardware of image processing platform, moreover, it researched and completed image algorithm on that platform. The system imaging on the same target scene with its image sensor array which composed by infrared imaging and low-light CCD camera. For the gray-scale image output by infrared camera is not suitable for human eyes to observe, the system preprocessed the original image at first, then segmented the image output by infrared camera into three regions of mountain, water and sky, and pseudo-color-coded the regions separately. It enhanced the target recognition rate of infrared image. For the image output by low-light CCD camera, the system finished the image algorithm which can reduce the mist impact. Finally, the system can output clear color images or videos throughout all-day by manual gating, provide help for the work of search and rescue and command and control.Experiments in local waters and on-site installation and testing for the system proved that the technical solution is feasible, the system completed the capture, processing, transmission and display for multi-band image. the images or videos it outputting met the basic requirements of the inland ship accident scene imaging. Currently, the system has been completed and accepted and put into use officially.