We present an automatic occlusion removal methodology for occluded images. The occlusion here considered are the images which contain elements such as grid or fence, the reflection of objects through glass windows and raindrop. The appearance of any object in the space which blocks the complete view of another object or a scene considers as the occlusion. Because of occlusion, we lost the aesthetic beauty of the desired scene. To obtain a background scene without any discrepancies occlusion removal is essential. Occlusion may happen accidentally, and also there are some situations we cannot avoid occlusion. For example, taking photos or videos in a zoo, fence removal is impossible. If the fence obstruction removes from photos, the results became awesome. The Aim is to improve the accuracy of occlusion removal system using an optical flow method. The sequence of frames considers as input to the system. The system automatically detects occlusion. The decomposition of the background component and occlusion components are done using an optical flow method. Finally estimates desired background scene while removing the annoying occlusion. We show results of experiments in the various occluded situation while taking photos or videos.

Occlusion, Optical flow, Detection, Decomposition

[1] Criminisi, Antonio, Patrik Perez, and Kentaro Toyama, “Region filling and object removal by exemplar-based image inpainting”, IEEE Transaction on image processing 13.9, pp.1200-1212, 2004
[2] Newson, A., Almansa, A., Fradet, M., Gousseau, Y., Perez, P., et al. “ Video inpainting of complex scenes.”, Journal on Imaging Sciences, Society for Industrial and Applied Mathematics, 2014
[3] Kao, Shannon, “Light Field Occlusion Removal”, Sanford University, 2005
[4] TianfanXue, Michael Rubinstein, Ce Liu, William T Freeman, “ Computational Approach for Obstruction -Free Photography”, ACM Transactions on Graphics, Vol. 34, No. 4, Article 79, 2015.
[5] Park, Minwooi et al. “Image de-fencing”, Asian Conference on Computer Vision. Springer, Berlin, Heidelberg, 2010
[6] Liuy, Yanxi, et al. “Image de-fencing”, Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on IEEE, 2008
[7] Jonna, Sankaraganesh, et al. “A multimodal approach for image de-fencing and depth inpainting”, Advances in Pattern Recognition (ICAPR), 2015 Eighth International Conference on. IEEE, 2015
[8] Yamashita, Atsushi, Akiyoshi Matsui, and Toru Kaneko. “Fence Removal from multi-focus images.”, Pattern Recognition (ICPR), 2010 20th International Conference on IEEE, 2010
[9] Garg, Kshitiz, and Shree K. Nayar., “Detection and removal of rain from videos”, Computer Vision and Pattern Recognition, 2004. CVPR 2004, Proceedings of 2004, IEEE Computer Society Conference on.Vol.1. IEEE, 2004
[10] Strecha, Christoph, RikFransens, and Luc Van Gool., “Wide-baseline stereo from multiple views: a probabilistic account.”, Computer Vision and Pattern Recognition, 2004. CVPR 2004, Proceedings of 2004, IEEE Computer Society Conference on.Vol. 1. IEEE, 2004
[11] Luan, Xiao, et al., “Extracting sparse error of robust PCA for face recognition in the presence of varying illumination and occlusion.”, Pattern Recognition, 47.2 PP. 495-508. 2004
[12] Ashraf Siddique and Seungkyu Lee., “Video Inpainting for Arbitrary Foreground Object Removal”, IEEE Winter Conference on Applications of Computer Vision, 2018
[13] VanshajSikri., “Proposition and Comprehensive Efficiency Evaluation of a Foreground Detection algorithm based on Optical Flow and Canny Edge Detection for Video Surveillance Systems “, IEEE WiSPNET conference. 2016
[14] T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. “High accuracy optical flow estimation based on a theory for warping” European Conference on Computer Vision (ECCV), 2004
[15] OpenCV Open source Computer Vision_optical Flow(https://docs.opencv.org/3.3.1/pages.html)