Parameters control of a conveyor belt camera system to improve vision inspection performance in industrial applications

This thesis is centered on the study and implementation of an algorithm designed to effectively control the parameters of Luxonis cameras in response to varying conditions. The primary objective is to enhance the performance of conveyor belt systems, particularly in industrial applications such as quality control and quantity estimation. Two distinct approaches, derived from two separate papers, were implemented, and a third paper was studied but not put into practice. The results obtained from these implementations are promising, however, incompatibilities with the cameras prevented comprehensive testing and calibration. In addition to the algorithmic implementations, the Robot Operating System (ROS) was incorporated to facilitate communication with other devices. The study also includes the calculation of the optimal speed of the conveyor belt, and the implementation of evaluation parameters for the images. These parameters serve as a metric for assessing the quality of the frames.