Color Particle Image Velocimetry Improved by Decomposition of RGB Distribution Integrated in Depth Direction

A single-camera color PIV system based on the combination of a color-coded illumination and a color high-speed video camera has been developed. The system realizes flow velocity vector fields in multiple layers in the depth direction for a relatively wide volume. The working principle belongs to smoke-imaging PIV since particle per pixel is larger than unity in the digital image plane, which is different from the case of color 3-D PTV. The accuracy of the present color PIV depends on how the integral equation of color-signal overlapping among three wavelengths (R,G,B) is solved as inverse analysis. We propose in this paper a method to reasonably solve the equation, which removes color contamination in RGB sensors, and decomposes the color-illuminated smoke images into the smoke distributions inside individual layers. The first half of the paper deals with linear backward analysis for the color de-contamination to establish triple-layer velocimetry by a single-view imaging of smoke. Based on the success, we present in the second half an extended system that realizes simultaneous six-layer color PIV to which three intermediate colors are additionally illuminated into the measurement volume.