Variation reduction in quality of an optical triangulation system employed for underwater range finding

This paper attempts to more thoroughly understand the factor effects on the performance of an optical triangulation system used for range finding in water. In this study analytical expressions for estimating the peak of the laser spot and for image processing were presented. Both the design of experiment process and the Taguchi method have been used to evaluate factor effects and to specify-design parameters. To assess the effects of monochrome and color image displays on the performance of range finding, two sets of experiments were carried out. The optimum models based on factor effects were predicted and constructed through the analysis of variance. The confirmation experiments were performed and the outcomes are in good agreement with the predicted optima. Based upon the results of experiments and analyses, it is found that brightness, contrast, and peak detection algorithms are the main factors to dominate the quality of range finding while using a monochrome camera. And further, all control factors of brightness, contrast, peak detection algorithm, sampling range, and specific primary color for intensity computation are critical to the quality of range finding while using a color camera. Notably, right selections of the image resolution and the sampling range can shorten the computing time and get high precision results. (C) 2002 Elsevier Science Ltd. All rights reserved.