A low light level image enhancement algorithm based on double branch pyramid model is proposed. It is used to solve the problems of low brightness, serious loss of detail, over exposure and color distortion of the low light level image enhanced by existing algorithms. Firstly, the low-light level image is converted from RGB color space to HSV color space. Secondly, the low-light level image enhancement network of double branching pyramid network is constructed for the V component, and the image features are obtained adaptively. The network structure is composed of two parallel branches. The branch with hierarchical residual module can effectively enhance the brightness of V component, and the branch with feature pyramid attention module can obtain deep feature information. Finally, the information extracted from the dual-branch structure is fused, and the enhanced image is converted from HSV color space to RGB color space. The peak signal-to-noise ratio and the average structure similarity of the proposed algorithm on real images are 29.451 dB and 0.930 1 respectively, which are higher than other comparison algorithms. Experimental results show that the enhancement of the V component can effectively improve the image brightness and restore the image details.