Image processing for spray analysis

R&D Vision has developed a number of algorithms for image processing in order to determine, in an objective and quantitative manner, certain characteristics of a spray (cone of the spray, length of penetration, wettability / impact). 

The use of a robot to generate controlled and reproducible sprays has helped us to implement several measurement diagnostics to compare and match results and validate the performance of the algorithms we have developed. To determine the angle of the cone of the spray, the results of tomographic measurements by laser beam plane (median plane) were compared to anintegral method using images of the spray obtained by a high speed camera (500 im/s) and synchronised LED lighting. The results from the two techniques are very similar thus validating an algorithm that no longer uses subjective thresholds to determine the angle of the cone of the spray. 

To determine the impact / wettability we generate a laser plane perpendicular to the direction of the spray at a configurable distance from the exit point of the nozzle. The drops that cross through the plane are imaged in forward motion by a high speed camera. The drops detected during spraying are accumulated on a 2D web to build a standard histogram. The spatial spread of the density of probability of the presence of drops is thus determined to be able to quantify and understand the shape and the proportion of liquid received by a surface at a certain distance from the nozzle.