Vehicle Tire Quality Mark Identification using Moving Window Based Approach

Vehicle Tyre inspection has been a crucial activity for identifying tyre manufacturing defects. Currently, the inspection process is manual and time-consuming. To make the inspection process quicker, simpler, and more accurate, with lower cost, it is essential to make the Tyre inspection process automatic. This automation is possible using advanced computer vision techniques. However, this technique also faces difficulties of low contrast image processing, as there is hardly any contrast difference between the quality marks and the tyre background. In this paper, we proposed a novel approach to overcome this black-on-black processing problem. Our new approach is based on a moving window-based approach, which allow to focus on a smaller area at a time to detect the presence of Quality marks like ISI or TISI on the tyre. Moving window-based approach to overcome the low contrast image processing problem and reduce the region of interest during image processing. The proposed method was modelled in Matlab and implemented on the Raspberry Pi platform. Marking speed and marking accuracy are improved. Processing speed is improved to 2 second per tyre as compared to 15 second manually. Implementation results show that the identification process is quick and more accurate. With the help of our approach, enterprises become more sustainable.