Research On Methods Of Far-infrared Enhancement And Fusion

Recent years, infrared images have been more and more applications in the military and civilian, infrared image processing-related technologies have also been great development, this paper studies the principle of the infrared imaging systems, infrared images, improving particle swarm optimization algorithm for infrared image enhancement as well as infrared and color image fusion algorithm based on wavelet transform.In this paper, with the standard optimization algorithm in the infrared image enhancement, there is a calculated slow, poor and premature convergence issues, so we proposed an improved particle swarm algorithm for image enhancement, improved particle swarm optimization algorithm is proposed based on the standard particle swarm incorporated herein a "multi-particle swarm" and "particle evolution" theory approach, using methods independent of each other to find the global optimal solution plurality of particles, effectively improve the global search capability PSO. Meanwhile, the particle swarm optimization iterative process, appropriate adjustments constantly accelerating factor inertia weight, both to avoid particle swarm into local optimal solution and cannot jump, particle swarm fast and easy to find in the later iterations of the global optimal solution. The infrared image processing simulation results show that the improved particle swarm image enhancement algorithms mentioned herein and standard particle swarm algorithm for image enhancement, compared with faster computing speed, better global convergence, with the common good and promote the value of.The image forming characteristic of the infrared image and color image has significant difference, but the fused images of them have good complementary information. This paper is based on the characteristics of the infrared image and color image, firstly Hue, Intensity, Saturation(HIS) color space conversion is used to extract the luminance component of the color image and wavelet transform is uses to the luminance component and the infrared image. Then the regional standard deviation fusion rule is adopted to the low frequency coefficient and the regional standard deviation ratio fusion rule is adopted to the high frequency coefficient and the image enhancement operator is applied on the image. Lastly the inverse wavelets transform and inverse HIS transform is used to obtain the finally fused image. The experimental results show that the fused image of this paper is better in the reservation of the useful information of the original infrared image and the color information of the original color image than other method.