Table 2 Comparisons of resource usages among different skin detector designs
From: Chip-based adaptive skin color detection using trajectory constraints on hue
Items | (1) | (2) | (3) | (4) | (5) | (6) |
|---|---|---|---|---|---|---|
Resources | Chenaoua and Bouridane ([2006]) | Fotouhi et al. ([2009]) | Khan et al. ([2011]) | Liu and Shi ([2011]) | Zafarifar et al. ([2012]) | Proposed TC |
Platforms | PC | PC | PC | FPGA | FPGA | FPGA |
EP2C20 | Xilinx | Cyclone IV E | ||||
EP4CE115F29C | ||||||
7Â N | ||||||
Methods | Markov random field | Texture analysis & ANN classifier | Cascaded skin training & classifier | YCbCr & I1I2I3 Classifiers | YUV & Adaptive texture analysis | Adaptive HSV constraints |
Tested images | Static picture | Static picture | Static picture | Video | Video | Video |
Adaptability | Yes | Yes | Yes | No | Yes | Yes |
Resource usages (LEs/LUTs) | - | - | - | 4,091 | 30,720 | 7,646 |
DSPs | - | - | - | No | 192 | No |
Embedded memory (bits) | - | - | - | 36,512 | 3,456,000 | 57,400 |