Hyperspectral image(HSI) has a wide range of application prospects because it contains rich image information. However, in a large amount of information, it usually contains a lot of noise, which will have a great impact on image processing and application. The biggest difficulty of hyperspectral image denoising lies in how to combine the information of many bands instead of denoising only one band. In this paper, a joint band hyperspectral image denoising model (HSI-UBDAE) based on the encoder is proposed, which makes full use of the correlation between bands and combines the spatial domain with the spectral domain to denoise HIS image effectively. Experiments on several real data sets show that the proposed method is superior to some mainstream denoising methods.