Images taken underwater will appear greenish or blueish due to the scattering and attenuation of light, resulting in color distortion of underwater images. This paper proposes a new underwater image defogging algorithm, which can effectively remove image noise, improve the color saturation and contrast of underwater images, and restore clear underwater images without fog. The algorithm introduces guided filtering, which increases speed and work efficiency. The linear enhancement of the saturation component can effectively enhance the color saturation of the underwater image. Compensation for local brightness can better show image details. Through experimental comparison with other algorithms, our algorithm has achieved good advantages in terms of defogging, image structural information integrity, and resistance to color attenuation.