Dynamic background subtraction with masked RPCA
- PDF / 950,769 Bytes
- 8 Pages / 595.276 x 790.866 pts Page_size
- 105 Downloads / 246 Views
ORIGINAL PAPER
Dynamic background subtraction with masked RPCA Hyomin Ahn1 · Myungjoo Kang1 Received: 6 February 2020 / Revised: 29 June 2020 / Accepted: 12 August 2020 © Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Robust principal component analysis (RPCA), a method used to decompose a matrix into the sum of a low-rank matrix and a sparse matrix, has been proven effective in modeling the static background of videos. However, because a dynamic background cannot be represented by a low-rank matrix, measures additional to the RPCA are required. In this paper, we propose masked RPCA to process backgrounds containing moving textures. First-order Markov random field is used to generate a mask that roughly labels moving objects and backgrounds. To estimate the background, the rank minimization process is then applied with the mask multiplied. During the iteration, the background rank increases as the object mask expands, and the weight of the rank constraint term decreases, which increases the accuracy of the background. We compared the proposed method with state-of-art, end-to-end methods to demonstrate its advantages. Keywords Dynamic background subtraction · RPCA · Markov random field
1 Introduction Background subtraction is a major approach to moving object detection and is required for automated video analysis and its practical applications in traffic monitoring, auto driving, and fire detection. The concept is to model a background without moving objects and compare it with each input video frame. Pixels with differences between them, which are greater than a set threshold, are considered moving objects. It is of significance that the background is not always static. That is, there is a difference in the relative importance of individual moving objects, which means that sometimes only important objects should be extracted. This problem is termed dynamic background subtraction. Robust principal component analysis (RPCA) [6] is a useful approach for static background subtraction owing to its ability to recover low-rank matrices. That is, the low-rank output of the RPCA models the static background, and the sparse output models the moving object. However, as the low-rank output of the RPCA is literally close to the low-
B
Myungjoo Kang [email protected] Hyomin Ahn [email protected]
1
Department of Mathematical Sciences, Seoul National University, 433, 27-433, 1, Gwanak-ro, Gwanak-gu, Seoul, Republic of Korea
rank matrix, its ability to express the dynamic background is limited. To compensate for this limitation, we introduced a mask to distinguish the object and background. We found that it supports modeling the dynamic background and selecting the major objects among the combination of objects. By combining the prevalent RPCA framework and the object mask, we produce another RPCA model for dynamic background subtraction. The contributions of this paper are as follows: • We propose a new type of RPCA-based dynamic background subtraction method that differently emphasizes the similarit
Data Loading...
