Image dehazing based on dark channel prior and brightness enhancement for agricultural remote sensing images from consumer-grade cameras

Authors: ZHANG, JIAWEI - Northwest Agricultural & Forestry University; WANG, XIUYUAN - Northwest Agricultural & Forestry University; Yang, Chenghai; JIAN, ZHANG - Northwest Agricultural & Forestry University; HE, DONGJIAN - Northwest Agricultural & Forestry University; SONG, HUAIBO - Northwest Agricultural & Forestry University

Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/20/2018
Publication Date: 12/29/2018
Citation: Zhang, J., Wang, X., Yang, C., Jian, Z., He, D., Song, H. 2018. Image dehazing based on dark channel prior and brightness enhancement for agricultural remote sensing images from consumer-grade cameras. Computers and Electronics in Agriculture. 151:196-206. https://doi.org/10.1016/j.compag.2018.06.010.
DOI: https://doi.org/10.1016/j.compag.2018.06.010

Interpretive Summary: The quality of remote sensing images is often affected by clouds, hazes and chaotic media in the atmosphere. Traditional image dehazing can enhance the brightness of the hazed image, but it can also cause brightness distortion. This study proposed an improved image dehazing approach based on a dehazing method and optimal selection of relevant enhancement parameters and evaluated the enhancement quality using four indicator indices. The results showed that dehazed images had improved quality compared with original images and that the proposed approach can effectively be applied to dehaze remote sensing images.

Technical Abstract: Remote sensing technology has been widely used for monitoring crop fields and other agricultural applications. However, the clarity of remote sensing images is often affected by clouds and chaotic media in the atmosphere. Image dehazing can be achieved through the dark channel prior method (DCP), but there is always a brightness distortion problem after image dehazing. To solve this problem, this study proposed an improved image dehazing approach based on the DCP method and optimal selection of relevant enhancement parameters. To verify the effectiveness of this approach, four evaluation indices including mean square error (MSE), peak signal to noise ratio (PSNR), average gradient, and program running time were used to evaluate the quality of enhanced images. By comparison, logarithmic enhancement was selected as the optimal enhancement method. Image enhancement achieved the best effect when the dark channel window size is 5, atmospheric light is 215/255, and the lower limit t0 of transmission factor is 0.1. Fifty airborne images from a consumer-grade camera flown by an agricultural aircraft were used to evaluate the improved method. Both the original and the enhanced images after dehazing were mosaicked by Adobe Photoshop software. The mosaicked images before and after image dehazing were compared. Results indicated that the mosaicked image without dehazing had an entropy of 6.359 and an average gradient of 6.513. In comparison, the mosaicked image with dehazing had an entropy was 6.668, 4.86% higher than the mosaicked image without dehazing, and an average gradient of 11.305, 73.58% higher than the original mosaicked image. These results indicate that the proposed method in this study can be applied to dehaze remote sensing images effectively.