Near-infrared Image Enhancement Method in IRFPA Based on Steerable Pyramid

Near-infrared imaging mainly uses near-infrared band ambient light imaging reflected by the target, which has better atmospheric penetration performance and human skin penetration performance than visible light imaging. Therefore, near-infrared imaging is widely used in military, medical and many industrial production. Aiming at reducing the noise and improving the contrast of the near-infrared images gained from infrared focal plane array (IRFPA), the near-infrared image enhancement method based on steerable pyramid is proposed in this paper. First of all, the near-infrared image is decomposed into multi-scales using the steerable pyramid model; then the coefficients of low-frequency and high-frequency of the image are obtained. In order to improve the contrast of the original near-infrared image, the coefficients with low-frequency are nonlinearly transformed through fuzzy-set theory. Then the coefficients of high-frequency are dealt with threshold method to reduce the noise. Next, these images are reconstructed. At last, anti-sharpening mask is used to highlight the details of the image. During the reconstruction, a adaptive interpolation algorithm is put forward to resolve the distortion problem in the steerable pyramid algorithm. The experimental results show that this algorithm has a good effect on the enhancement of near-infrared images, and significantly improves the quality of near-infrared image produced by IRFPA device. The comparison results with various algorithms show that our algorithm outperforms the state-of-art in terms of contrast gain, comentropy, mean-square error, and peak signal to noise ratio. The experimental results illustrate that the proposed algorithm can efficiently enhance the near-infrared image.