Event Detection Module for Low-Power Camera
- 1.6k Downloads
In this chapter, we first propose an effective low-power image sensor system for event detection. The system consisting of a low-resolution auxiliary sensor and a high-resolution main sensor operates in two different modes, sleep and wakeup. In the sleep mode, only the auxiliary sensor works for event detection and the main sensor remains off for power saving. In the wake-up mode, the main sensor turns on based on the data sensed by the auxiliary sensor and into normal operation. Second, a new background subtraction algorithm which can be used for event detection is proposed. The algorithm works much faster than conventional algorithms and requires less computation which is critical for low-power operation. In addition, utilization of depth information in background subtraction for 3-D image applications is also presented. Finally, hardware implementation of a low power low-resolution CMOS image sensor (CIS) is presented. The CIS fabricated in 0.18 μm CIS process is designed to be used as an auxiliary sensor in the proposed system. The CIS generates 4-bit image data and consumes only 1.45 mW out of 3 V supply.
KeywordsLow-power image sensor Low-power event detection Background subtraction Depth information CMOS FPGA
This work is supported by the Center for Integrated Smart Sensors funded by the Ministry of Science, ICT and Future Planning as the Global Frontier Project.
- 2.Fossum ER (1997) CMOS image sensors: electronic camera-on-a-chip. IEEE Trans Electron Devices 44(10):1689–1698Google Scholar
- 4.Cucchiara R, Piccardi M, Prati (2003) A detecting moving objects, ghosts, and shadows in video streams. IEEE Trans Pattern Anal Mach Intell 24:1337–1342Google Scholar
- 8.Yao J, Odobez J-M (2007) Multi-layer background subtraction based on color and texture. In: IEEE conference on computer vision and pattern recognition, June 2007Google Scholar
- 9.Noh SJ, Jeon M (2012) A new framework for background subtraction using multiple cues. In: The 10th Asian conference on computer visionGoogle Scholar
- 10.Harville M, Gordon G, Woodfill J (2001) Foreground segmentation using adaptive mixture models in color and depth. In: Proceedings of the IEEE workshop on detection and recognition of events in video, IEEE computer society, Los Alamitos, CA, USA, pp 311Google Scholar
- 12.Fernandez-Sanchez EJ, Rubio L, Diaz J, Ros E (2013) Background subtraction model based on color and depth cues. Mach Vis Appl 25(5):12111225Google Scholar
- 13.Stauffer C, Grimson W (1999) Adaptive background mixture models for real-time tracking. In: Proceedings of IEEE conference on computer vision and pattern recognition (CVPR 1999), pp 246–252Google Scholar
- 15.Sobral A, Bender L, Parks D, Yao J, Odobez J-M, Noh SJ “A background subtraction library” in GitHub. https://github.com/andrewssobral/bgslibrary