Design And Development Of Night Vision Helmet Based On Low Light And Infrared Fusion

With the rapid development of night vision technology, low light level and infrared imaging detectors are widely applied in social security monitoring. However, since single band imaging sensor is not able to provide comprehensive information, or not conducive to people to capture important information, it is necessary to fuse multiband image information to implement accurate monitoring of social security for police officers. According to the requirements of the safety monitoring system, including small size, light weight and portability, this paper designs a new type of night vision helmet, which can realize the low light visibility and infrared image fusion.In order to overcome the shortcomings of security monitoring system, the paper conducts the research and design of night vision helmet system. In the system, the source images are captured by WAT-902H low light level camera and GT320CM-A00 infrared camera which are known for small size and light weight, and DM642 is used to accomplish the image fusion. Finally, the fused image is displayed in PAL format for visual observation.Image fusion processing is the important part of the night vision helmet system. As a result, based on the comparative research of the existing image fusion algorithms, two new fusion algorithms of high-low frequency fusion rules are proposed in the article, to improve the fused image with low texture. First is the local feature residuals and statistical matching based image fusion algorithm. Second is the local gradient correlation and statistic feature based image fusion algorithm. These two algorithms are proved to extract the texture information sufficiently and efficiently by applying the analysis of wavelet direction and image local statistical features. According to the calculated performance of DM642 and the fusion algorithms, the existing algorithms are optimized, which further removes the difficulty of real-time image fusion and display.In conclusion, the paper verifies the fusion and color display functions of the night vision helmet comprehensively. After experiments under different circumstances, the improved algorithms are proved to be effective. Moreover, the integrated system is qualified to provide a good real-time performance with rich fusion image, which is believed to assist the police to detect the suspicious targets in real time.