Research On Night Vision On Vehicle Base On Infrared Camera

As you know, lighting in the road is close in night, the driver can not be timely master the road conditions, which are the most important reasons for traffic accidents at night. Therefore, this paper developed the infrared thermal imaging night vision system suited for vehicle. The thermal imaging technique is based on the temperature difference between the target and the background. It overcomes that valid distance is short and night vision system is vulnerable to the vehicle headlight interference. It greatly improves driving safety at night and special weather. It has the very high practical value.First of all, the paper elaborates the night vision system’s important significance on improving the car’s active safety and introduces the international research situation on night vision system. Secondly, according to the basic requirements of night vision, the paper gives the design of the whole system and module partition, mainly including the focal plane array detector module, image acquisition and processing module, TEC module, image display module. The infrared thermal imaging system’s modules are introduced in detail from the module design of hardware circuit and the software program, and giving the theoretical analysis and calculation.However, currently the imaging quality is always not high enough, because of the limitation of infrared detector. IRFPA device has Non-uniformity correction and blind pixel. This paper makes a comparison among several kinds of non-uniformity correction algorithms, blind pixel compensation algorithms and image enhancement algorithms, and chooses the suitable algorithm for the system hardware platform. This system uses two-point temperature Non-uniformity correction algorithm to solve image Non-uniformity problem. The neighborhood pixel method applies to blind pixel compensation. At the same time we use the median filter to highly improve image quality.At last, we do some experiments at lab and load at night. Experimental data shows that the target image has good effect and strong contrast, which the people and vehicles can be identified in. Thermal imaging night vision system is active to meet the design requirements.