Research On Infrared Image Enhancement Technology

Nowadays infrared imaging technology has a broad application prospect. However, the imperfections of great non-uniformity, low contrast and poor resolution remain widespread in infrared images which limit the development of infrared imaging technology. Therefore we do depth research on infrared image enhancement technology in three aspects as non-uniformity correction, contrast enhancement and detail enhancement.For non-uniformity correction, first, this paper proposes the non-uniformity correction technology based on A/D drive circuit and the on-chip non-uniformity correction technology based on matching calibration words of the infrared detector which not only reduce the non-uniformity of infrared images, but also improve the dynamic range of the infrared imaging system. Second, we put forward a non- uniformity correction technology with no baffle, improving the stability and reliability of the infrared imaging system. Finally, we come up with an improved non-uniformity correction technology based on SURF image registration which has improved the applicability of rotating and scaling image sequences.For contrast enhancement, we put forward the adaptive grayscale adjustment-based stripe noise removal method and dual-platform histogram equalization of adaptive dynamic range. The former takes advantage of stripe noise spectrum to compute column histogram, so that to remove stripe. The latter sets a threshold of stretched grayscale image dynamic range by counting valid grayscale on the basis of dual-platform histogram equalization algorithm. And they all have substantially improved the infrared image contrast.For detail enhancement, this paper presents a BF & DRP image detail enhancement technology based on multi-layer separation. This method separates the infrared image that after denoising into multiple layers. Furthermore we make different enhancement on these hierarchical images. The final image is superimposed by all processed hierarchical images. This method not only enhances the image detail, and can eliminate pseudomorphism that exist in former detail enhancement techniques, but also increases the signal to noise ratio of the image.Experimental results show that the proposed image enhancement algorithm can effectively reduce the infrared image noise and improve the contrast of infrared image, which has great application value of improving the infrared image quality.