Skip to main content
. 2012 Oct 22;12(10):14129–14143. doi: 10.3390/s121014129

Table 1.

Segmentation algorithm, time required to perform the operations and the real frame rate achieved.

Algorithm Time (ms) fps
Image copy from the FIFO to processor memory 40.30 10
Image reading from the FIFO +
Threshold segmentation of 1 color layer [example: R > 16] +
Result stored into processor memory
44.28 10
Image reading from the FIFO +
RGB threshold segmentation [example: R > 16 & G < 55 & B < 12] +
Result stored into processor memory
46.61 10
Image reading from the FIFO +
RGB interval segmentation [ex.: 32 > R > 12 & 55 > G > 10 & 16 > B > 12] +
Result stored into processor memory
52.23 10
Image reading from the FIFO +
64 KB LUT segmentation using the RGB565 16 bits value as an index [ex.: s = LUT(RGB565)] +
Result stored into processor memory
44.41 10
Image reading from the FIFO +
8 KB LUT segmentation using 13 bits as a pointer and 3 bits as byte offset [ex.: s = LUT(RGB565(1:3), RGB565(4:16),)] +
Result stored into processor memory
51.73 10
Image reading from the FIFO +
8 KB LUT segmentation using 13 bits as a pointer and 3 bits as byte offset [ex.: s = LUT(RGB565(1:3), RGB565(4:16),)] +
Result stored into processor memory +
Red Peach centroid computation
63.64 10
Image reading from the FIFO +
Linear Color Model classification and segmentation (11 classes) +
Result stored into processor memory
2,228.47 0.44