In thick fog conditions,the medium with uneven density will cause light scattering,reflection and attenuation in light transmission process,resulting in serious image degradation,so there is difficulty in obtaining image information from UAV aerial photography.In order to solve this problem,a bilateral-filtering transmittance optimization algorithm based on boundary constraints is proposed for image defogging.Firstly,the colors of the image are corrected to enhance the brightness of the image.Secondly,according to the inherent boundary constraints of the transmission function,the boundary constraints are constructed and then combined with context regularization,and the initial transmittance of the image is obtained and then the transmittance is refined through repeated calculation.Then,the refined transmittance is Gaussian-weighted to minimize noise interference,and the optimal transmittance is obtained by bilateral filter.Finally, the global atmospheric light of the image is evaluated and the dark channel theory is used to defog the image,so that a sharpened image is obtained.Based on subjective visual effects and objective experimental data analysis,it is proved that the proposed method can effectively remove the fogging phenomenon in degraded images,and has a significant advantage over the existing novel image defogging algorithms.