The Infrared Small Target Detection Based On Wavelet Transform Technology Research

In the field of computer vision, radar tracking, infrared guidance and television tracking, target detection is an extremely important position. In practical applications, especially in military applications, the distance between target and sensors is very large, and the background and environment of the targets are more complex. The target can be easily submerged in the noise inside in the condition more noise. In this case, some of the more traditional image processing, target detection does not seem to have an advantage. Quickly and high degree of accuracy target detection technology has become a favorite in the field of image processing.The focus of this article is on small target detection based on wavelet transform. The first analysis the characteristics of wavelet transform, and analysis of the advantages of wavelet transform to other transforms. Mainly reflect the multiresolution wavelet transform and time-frequency window on these two points. The wavelet transform can decompose the signal into different frequency bands, in general, aim at high frequencies and the background is mainly in the low frequency band, this decomposition can effectively remove the background.First, I do some research of the wavelet transform theoretical knowledge in this paper, and analyze its characteristics.Secondly, the image pre-processing based on wavelet transform, and a filter better, more sophisticated noise suppression filter. Also introduced other image filtering method, the experiments available, this method can better noise suppression to retain the details of the objectives.Again, carefully studied the edge detection method based on wavelet transform analysis and experimental results of several wavelet-based edge detection method.Finally, Introduction two small target detection method based on wavelet transform, a simple threshold segmentation detection method and energy-crossing detection method based on wavelet transform, analysis and the experimental results, obtained energy-crossing detection method based on wavelet transform has better performance indicators.Finally, I analyze this study and future outlook.