This technology is a multi-sensor interpolation device and method that calculates horizontal and vertical green evaluation values at R or B pixel positions to obtain an average directional evaluation, and applies weights for each direction—derived from the difference between surrounding green values and the red average—to horizontal and vertical interpolation values to obtain the final green value.
Mosaic images captured by color CCD/CMOS sensors in digital cameras and camcorders require interpolation because two colors are missing at each pixel. Existing methods often focus on either the horizontal or vertical axis, failing to adequately compensate for distortions such as zipper effects, color moiré, detail blurring, and false colors.
This technology calculates a directional evaluation by adding the average difference of adjacent red (blue) values to the average of left/right and top/bottom green values. It then creates horizontal and vertical weights using values from left, right, top, and bottom weight generators to perform a weighted average of the two directional interpolation values. By simultaneously reflecting vertical and horizontal axis components, it reduces distortions caused by single-direction judgment errors. Applicable to various CFA sensor-based equipment like camcorders, security cameras, and automotive cameras, it suppresses zipper effects and false colors with a compact hardware structure.
N/A
