TY - GEN
T1 - Image enhancement of shadow region based on polarization imaging
AU - Reda, Mohamed
AU - Shen, Linghao
AU - Zhao, Yongqiang
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019
Y1 - 2019
N2 - Object of shadow area suffers from low contrast and noise, causing difficulties for object detection and recognition. However, conventional image enhancement methods perform inefficiently in severer shadow conditions, where the contrast enhancement is not obvious, or the noise amplification is huge. In fact, Polarization information can be preserved even in very heavy shadows. Combining the polarization information with the intensity and chromaticity information of the original image can overcome the above problems. This paper proposes a method of object enhancement in shadow based on polarization imaging. Firstly, this paper puts forward degree of polarization (DOP) Invariance Prior and DOP-Chromatic Consistency Prior. Then, based on these two priors, a feedback image enhancement method based on polarization imaging is proposed. Experiments show that the proposed method is superior to the other conventional image enhancement methods.
AB - Object of shadow area suffers from low contrast and noise, causing difficulties for object detection and recognition. However, conventional image enhancement methods perform inefficiently in severer shadow conditions, where the contrast enhancement is not obvious, or the noise amplification is huge. In fact, Polarization information can be preserved even in very heavy shadows. Combining the polarization information with the intensity and chromaticity information of the original image can overcome the above problems. This paper proposes a method of object enhancement in shadow based on polarization imaging. Firstly, this paper puts forward degree of polarization (DOP) Invariance Prior and DOP-Chromatic Consistency Prior. Then, based on these two priors, a feedback image enhancement method based on polarization imaging is proposed. Experiments show that the proposed method is superior to the other conventional image enhancement methods.
KW - DOP Invariance Prior
KW - DOP-Chromatic Consistency Prior
KW - Polarized imaging
KW - Shadow enhancement
UR - http://www.scopus.com/inward/record.url?scp=85076949731&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-31723-2_63
DO - 10.1007/978-3-030-31723-2_63
M3 - 会议稿件
AN - SCOPUS:85076949731
SN - 9783030317225
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 736
EP - 748
BT - Pattern Recognition and Computer Vision 2nd Chinese Conference, PRCV 2019, Proceedings, Part II
A2 - Lin, Zhouchen
A2 - Wang, Liang
A2 - Tan, Tieniu
A2 - Yang, Jian
A2 - Shi, Guangming
A2 - Zheng, Nanning
A2 - Chen, Xilin
A2 - Zhang, Yanning
PB - Springer
T2 - 2nd Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2019
Y2 - 8 November 2019 through 11 November 2019
ER -