Design of Infrared Small Target Detection Software Based on Energy Minimization Constraints

In the complicated high sky environment, when the signal noise ratio (SNR) is low, the infrared small target which is collected by the infrared detection instrument is difficult to be detected. The infrared weak and small target detection is researched. On the basis of noise character analysis of infrared image, an improved infrared small target detection algorithm is proposed based on energy minimization constrained de-noising technology. Firstly, the image energy model is established, the energy minimization constraint technology is used to make the average output energy of local image area be minimum, and the infrared image noise and clutter signal are suppressed. The directional gradient features of infrared target are extracted, and the potential small targets can be identified with its direction gradient in sequential frames. The multiple frames determination results are obtained, and the fusion result is used to detect the target. Simulation result shows that the algorithm does not need the prediction of background, and it can overcome the interference of cloud in high altitude sky with low SNR, the small target detection accuracy is improved greatly. The algorithm has good application value in target detection.