A 0.18 μm CMOS on-chip skin detection scheme based on NPNP-triple-junction structure

In this paper, a 0.18 μm CMOS color processing scheme based on NPNP-triple-junction structure is proposed to achieve real-time skin detection. The need for building complex demosaicing circuitry is completely eliminated, which dramatically reduces the fabrication cost and simplifies the pixel space sampling algorithm. In addition, real-time detection is enabled through the use of logarithmic readout mode and Gilbert normalization cell. The fabricated chip and related measurement results have verified the concept of a compact, low cost, CMOS fully compatible micro-sensor suitable for skin detection.

[1]  A. El Gamal,et al.  CMOS image sensors , 2005, IEEE Circuits and Devices Magazine.

[2]  Christof Koch,et al.  Toward color image segmentation in analog VLSI: Algorithm and hardware , 1994, International Journal of Computer Vision.

[3]  E. Culurciello,et al.  A biomorphic digital image sensor , 2003, IEEE J. Solid State Circuits.

[4]  R. Etienne-Cummings,et al.  Color segmentation, histogramming and pattern matching chip , 2002, 2002 IEEE International Symposium on Circuits and Systems. Proceedings (Cat. No.02CH37353).

[5]  King Ngi Ngan,et al.  Face segmentation using skin-color map in videophone applications , 1999, IEEE Trans. Circuits Syst. Video Technol..