Retinex based visual identicalness detection for videos corrupted by imaging noise
作者：
Highlights：
• A new approach to detect visual identicalness for the noisy videos using its Retinex image to eliminate noise influence.
• Robust to heterogeneous video contents and cameras with much higher detection rate in comparison with de-noising approaches.
• High storage and memory efficiency compared with the existing background subtraction approaches [8], [9], [10].
• Efficient in reducing computational complexity and power consumption for real-time applications using the videos captured by low-cost cameras as input.
摘要
•A new approach to detect visual identicalness for the noisy videos using its Retinex image to eliminate noise influence.•Robust to heterogeneous video contents and cameras with much higher detection rate in comparison with de-noising approaches.•High storage and memory efficiency compared with the existing background subtraction approaches [8], [9], [10].•Efficient in reducing computational complexity and power consumption for real-time applications using the videos captured by low-cost cameras as input.
论文关键词：Noise robust technique,Identicalness detection,Retinex theory,Video applications
论文评审过程：Received 11 January 2013, Revised 23 May 2013, Accepted 27 June 2013, Available online 18 July 2013.
论文官网地址：https://doi.org/10.1016/j.image.2013.06.004
