Image Dehazing Using LiDAR Generated Grayscale Depth Prior

Sensors - Tập 22 Số 3 - Trang 1199
Won Young Chung1, Sun Young Kim2, Chang Ho Kang3
1Department of Aerospace Engineering, Automation and System Research Institute, Seoul National University, Seoul 08826, Korea
2School of Mechanical Convergence System Engineering, Kunsan National University, Gunsan 54150, Korea
3Department of Mechanical System Engineering (Department of Aeronautics, Mechanical and Electronic Convergence Engineering), Kumoh National Institute of Technology, Gumi 39177, Korea

Tóm tắt

In this paper, the dehazing algorithm is proposed using a one-channel grayscale depth image generated from a LiDAR point cloud 2D projection image. In depth image-based dehazing, the estimation of the scattering coefficient is the most important. Since scattering coefficients are used to estimate the transmission image for dehazing, the optimal coefficients for effective dehazing must be obtained depending on the level of haze generation. Thus, we estimated the optimal scattering coefficient for 100 synthetic haze images and represented the distribution between the optimal scattering coefficient and dark channels. Moreover, through linear regression of the aforementioned distribution, the equation between scattering coefficients and dark channels was estimated, enabling the estimation of appropriate scattering coefficient. Transmission image for dehazing is defined with a scattering coefficient and a grayscale depth image, obtained from LiDAR 2D projection. Finally, dehazing is performed based on the atmospheric scattering model through the defined atmospheric light and transmission image. The proposed method was quantitatively and qualitatively analyzed through simulation and image quality parameters. Qualitative analysis was conducted through YOLO v3 and quantitative analysis was conducted through MSE, PSNR, SSIM, etc. In quantitative analysis, SSIM showed an average performance improvement of 24%.

Từ khóa


Tài liệu tham khảo

Taketomi, 2017, Visual SLAM algorithms: A survey from 2010 to 2016, IPSJ Trans. Comput. Vis. Appl., 9, 16, 10.1186/s41074-017-0027-2

Shao, 2013, Feature learning for image classification via multiobjective genetic programming, IEEE Trans. Neural Netw. Learn. Syst., 25, 1359, 10.1109/TNNLS.2013.2293418

Zhu, 2014, Weakly-supervised cross-domain dictionary learning for visual recognition, Int. J. Comput. Vis., 109, 42, 10.1007/s11263-014-0703-y

Luo, 2014, Decomposition-based transfer distance metric learning for image classification, IEEE Trans. Image Process., 23, 3789, 10.1109/TIP.2014.2332398

Tao, 2008, Geometric mean for subspace selection, IEEE Trans. Pattern Anal. Mach. Intell., 31, 260

Liu, L., and Shao, L. (2013, January 3–9). Learning discriminative representations from RGB-D video data. Proceedings of the Twenty-Third International Joint Conference on Artificial Intelligence, Beijing, China.

Tao, 2007, General tensor discriminant analysis and gabor features for gait recognition, IEEE Trans. Pattern Anal. Mach. Intell., 29, 1700, 10.1109/TPAMI.2007.1096

Zhang, 2012, Slow feature analysis for human action recognition, IEEE Trans. Pattern Anal. Mach. Intell., 34, 436, 10.1109/TPAMI.2011.157

Wu, 2021, Multi-target recognition of bananas and automatic positioning for the inflorescence axis cutting point, Front. Plant Sci., 12, 705021, 10.3389/fpls.2021.705021

Gong, L., and Fan, S. (2022). A CNN-Based Method for Counting Grains within a Panicle. Machines, 10.

Forster, C., Pizzoli, M., and Scaramuzza, D. (June, January 31). SVO: Fast semi-direct monocular visual odometry. Proceedings of the 2014 IEEE International Conference on Robotics and Automation (ICRA), Hong Kong, China.

Engel, 2017, Direct sparse odometry, IEEE Trans. Pattern Anal. Mach. Intell., 40, 611, 10.1109/TPAMI.2017.2658577

Han, 2014, Object detection in optical remote sensing images based on weakly supervised learning and high-level feature learning, IEEE Trans. Geosci. Remote Sens., 53, 3325, 10.1109/TGRS.2014.2374218

Cheng, 2013, Object detection in remote sensing imagery using a discriminatively trained mixture model, ISPRS J. Photogramm. Remote Sens., 85, 32, 10.1016/j.isprsjprs.2013.08.001

Han, 2014, Efficient, simultaneous detection of multi-class geospatial targets based on visual saliency modeling and discriminative learning of sparse coding, ISPRS J. Photogramm. Remote Sens., 89, 37, 10.1016/j.isprsjprs.2013.12.011

2017, Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras, IEEE Trans. Robot., 33, 1255, 10.1109/TRO.2017.2705103

Qin, 2018, Vins-mono: A robust and versatile monocular visual-inertial state estimator, IEEE Trans. Robot., 34, 1004, 10.1109/TRO.2018.2853729

Zhang, J., and Singh, S. (2014, January 13–16). LOAM: Lidar Odometry and Mapping in Real-time. Proceedings of the Robotics: Science and Systems, Online Conference.

Shan, T., and Englot, B. (2018, January 1–5). Lego-loam: Lightweight and ground-optimized lidar odometry and mapping on variable terrain. Proceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, Spain.

Zhang, J., and Singh, S. (2015, January 26–30). Visual-lidar odometry and mapping: Low-drift, robust, and fast. Proceedings of the 2015 IEEE International Conference on Robotics and Automation (ICRA), Seattle, WA, USA.

Agarwal, A., Maturana, D., and Scherer, S. (2014). Visual Odometry in Smoke Occluded Environments, Robotics Institute, Carnegie Mellon University.

Narasimhan, 2002, Vision and the atmosphere, Int. J. Comput. Vis., 48, 233, 10.1023/A:1016328200723

Nayar, S.K., and Narasimhan, S.G. (1999, January 20–27). Vision in bad weather. Proceedings of the Seventh IEEE International Conference on Computer Vision, Kerkyra, Greece.

Narasimhan, S.G., and Nayar, S.K. (2001, January 8–14). Removing weather effects from monochrome images. Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2001), Kauai, HI, USA.

Zhu, 2015, A fast single image haze removal algorithm using color attenuation prior, IEEE Trans. Image Process., 24, 3522, 10.1109/TIP.2015.2446191

He, 2010, Single image haze removal using dark channel prior, IEEE Trans. Pattern Anal. Mach. Intell., 33, 2341

Zhao, 2020, Monocular depth estimation based on deep learning: An overview, Sci. China Technol. Sci., 63, 1612, 10.1007/s11431-020-1582-8

Tan, R.T. (2008, January 23–28). Visibility in bad weather from a single image. Proceedings of the 2008 IEEE Conference on Computer Vision and Pattern Recognition, Anchorage, AK, USA.

Fattal, 2014, Dehazing using color-lines, ACM Trans. Graph. (TOG), 34, 13, 10.1145/2651362

Huo, 2020, Fast Fusion-Based Dehazing with Histogram Modification and Improved Atmospheric Illumination Prior, IEEE Sens. J., 21, 5259, 10.1109/JSEN.2020.3033713

Zheng, 2020, Image dehazing by an artificial image fusion method based on adaptive structure decomposition, IEEE Sens. J., 20, 8062, 10.1109/JSEN.2020.2981719

Cai, 2016, Dehazenet: An end-to-end system for single image haze removal, IEEE Trans. Image Process., 25, 5187, 10.1109/TIP.2016.2598681

Ren, W., Liu, S., Zhang, H., Pan, J., Cao, X., and Yang, M.-H. (2016, January 11–14). Single image dehazing via multi-scale convolutional neural networks. Proceedings of the European Conference on Computer Vision, Amsterdam, The Netherlands.

Li, B., Peng, X., Wang, Z., Xu, J., and Feng, D. (2017, January 22–29). Aod-net: All-in-one dehazing network. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.

Li, R., Pan, J., Li, Z., and Tang, J. (2018, January 18–23). Single image dehazing via conditional generative adversarial network. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.

Li, 2020, Task-oriented network for image dehazing, IEEE Trans. Image Process., 29, 6523, 10.1109/TIP.2020.2991509

Zhao, 2021, RefineDNet: A weakly supervised refinement framework for single image dehazing, IEEE Trans. Image Process., 30, 3391, 10.1109/TIP.2021.3060873

He, 2012, Guided image filtering, IEEE Trans. Pattern Anal. Mach. Intell., 35, 1397, 10.1109/TPAMI.2012.213

Geiger, 2013, Vision meets robotics: The kitti dataset, Int. J. Robot. Res., 32, 1231, 10.1177/0278364913491297

Godard, C., Mac Aodha, O., Firman, M., and Brostow, G.J. (2019, January 27–28). Digging into self-supervised monocular depth estimation. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea.

Wang, 2004, Image quality assessment: From error visibility to structural similarity, IEEE Trans. Image Process., 13, 600, 10.1109/TIP.2003.819861

Bijelic, M., Gruber, T., Mannan, F., Kraus, F., Ritter, W., Dietmayer, K., and Heide, F. (2020, January 14–19). Seeing through fog without seeing fog: Deep multimodal sensor fusion in unseen adverse weather. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.