Target detection based on wavelet transform

Infrared image target detection problem is studied. According to the characteristics of infrared image noise distribution, a new denoising algorithm is proposed. The wavelet image is decomposed and reconstructed twice, and the iterative selection threshold strategy is applied to the wavelet coefficients after decomposition to eliminate noise. The simulation results show that the above algorithms are superior to the traditional denoising algorithm for both visual effects and objective evaluation indicators. And can be widely used in the field of infrared imaging. In terms of image enhancement, an image enhancement algorithm based on weighted adaptive local contrast is adopted for infrared image contrast and low signal-to-noise ratio. It takes into account both the enhancement of image detail and the suppression of noise. For the case where there is contrast difference in different areas of the image, the high-contrast area details are less enhanced by the parameter setting, and the low-contrast area is increased to enhance the detail, thereby improving the image visual effect. Since the calculation corresponding to the pixel whose gradient is smaller than a certain threshold value in the original image is neglected during the enhancement processing of the image, the amount of calculation required for image enhancement is reduced. For the segmentation of infrared images, an iterative selection threshold segmentation algorithm is adopted. The basic idea of the algorithm is to start selecting a threshold as the initial estimate and then continually update this estimate according to certain rules until the given condition is met. Compared with traditional threshold segmentation, the algorithm can segment the target region more accurately from complex backgrounds.