We analyze the current Bridgestone Matlab software for the detection of cracks in tires in order to suggest possible improvements to the software on the basis of the results of the analysis.

Our goal is to increase the perfomance, such as the recognition accuracy, to reduce number of false positives detected and to optimize the execution of the software, which is executed in online mode, reducing the time of the first detection of the cracks.

Federico Magliani is the contact for this project.