Abstract
In a diesel engine any malfunction of the injector degrades engine performance. Currently, the testing of injectors is done using Volumetric Testing, Illuminated spray test tube with high intensity LED or ITU-101 Common rail tester. The current methods require experienced and skilled personnel and are manual in nature. In this research, a novel computer vision based Injector Spray pattern Testing (VIST) system is designed to automate the testing of the diesel spray pattern using images acquired using a CCD camera. The macroscopic and microscopic features of the images are extracted, and analyzed for testing the reliability of the diesel injector. An interactive user interface is developed to display the results. As compared to the manual volumetric testing, the VIST system is more accurate because it uses individual plume features to distinguish between good and defective injector.