Due to uneven lighting, unknown suspended particle deposition, and rapid movement of cells in seawater, harmful algal bloom (HABs) micro-images collected in real-time marine engineering suffer from a series of issues such as poor clarity, blurred cell contours, and distorted textures. To solve these issues, this paper introduces an image enhancement method, which is based on image blurriness and multi-channel color compensation, and is developed using the Underwater Image Blurriness and Light Absorption approach. Considering the critical role of depth estimation in transmission estimation and image restoration, we first enhance the image by Gaussian blurring and proportionally merging it with the original image to improve edges and details, thus resulting in a sharpened image. Then, we utilize the relative global histogram stretching (RGHS) operation on white-balanced images to enhance image contrast, significantly improving the accuracy of depth and transmission estimation. Furthermore, we estimate the three color channels and use an appropriate transmission map estimation method. For instance, if the red channel is missing, indicating a bias towards cyan and green colors will estimate the missing components using the red channel, ensuring the method remains applicable to other color biases. We incorporate adaptive estimation of image brightness and darkness to achieve a more reasonable distribution of image brightness in image enhancement, avoiding excessively bright or dim results. Through experiments, our method’s enhancement of algal cell features surpasses that of other methods.