Composite Restoration of Infrared Image Based on Adaptive Threshold Multi-parameter Wavelet.

The multiplicative speckle noise and additive background noise of an infrared image are significant elements impacting image quality. To address the issue of image degradation caused by noise superposition and enhance the infrared image quality in terms of noise suppression, a composite restoration method based on adaptive threshold multi-parameter wavelet is proposed. First, based on the noise distribution characteristics of the infrared image, the multiplicative noise in the infrared image is transformed into additive noise, and the image is restored using the wavelet transform coefficient of the converted infrared image. Then, the benefits and drawbacks of soft and hard threshold functions are analysed in depth, and an adaptive double threshold function with adjustable parameters is developed. Finally, a fast non-local means method is used to suppress the effect of background noise on image quality. The experimental results show that the proposed method reduces 111.03 dB on average over the MSE index, 6.67 dB on the PSNR index and 6.92 dB on the SNR index.