Images are corrupted by various types of noise during acquisition and transmission. Designing a denoising algorithm that effectively removes noise while preserving the original edges, textures, and other fine details of the image has long been a goal of researchers. Based on the traditional wavelet thresholding image denoising algorithm, this paper proposes an improved image denoising algorithm that combines wavelet transform and total variation models. First, considering the correlation between wavelet coefficients at different levels in the discrete wavelet transform domain of the image, this method selects different adaptive thresholds for the high frequency subbands at each level and adaptively adjusts the shrinkage of the wavelet coefficients to remove high frequency noise from the image. Then, the total variation model is applied to remove low frequency noise while effectively controlling image distortion and smoothing. Experimental results demonstrate that this algorithm effectively removes noise while better preserving image features such as edges and textures. It outperforms other related denoising algorithms in terms of peak signal-to-noise ratio (PSNR), structural similarity(SSIM), and visual quality.