Study On Enhancement Algorithm Of Infrared Image

Thermal imaging systems are used in a wide range of applications both in military and civilian sectors along with double-quick development of modern infrared technology. High background, low contrast ratio, narrow gray-level range and low Signal Noise Ratio (SNR) are the characteristics of infrared images because of the effect of infrared imaging principle, thermal imaging systems and environment. The quality of infrared image is strongly affected by above factors, consequently, making the target of image vaguely and the result unideally. In practical application, the obtained infrared image must be enhanced to easy observe for improving the quality of image. So, it is key step to enhance infrared image in the technology of infrared image process.In this paper, the enhancement algorithm of infrared image is studyed through heightening SNR of the infrared image, restraining noise and standing out puny target based on the characteristics of infrared images and the conventional Histogram equalization (HE). The main work is:The characteristic of image histogram and noise are analysed respectively. The original infrared images are applied with the various methods to contrast the effect by simulation.Aimed at the the phenomenon of gray-level truncation and losing details, a new method of infrared image histogram equalization with image details preservation is proposed. It is the key to choose the threshold in the plateau histogram equalization (PHE), in this paper, a self-adaptable enhancing algorithm is present based on PHE. To make conspicuous the details of the image target, a type of non-linear and self-adaptable enhancing algorithm based on histogram non-linear extension is proposed.Theories and simulation experiment proves that the proposed algorithms in this paper turn to be concise and self-adaptable and can be used for handling hardware real-time processing, bring about a better enhancement to the infrared images.