Intelligent Speed Regulation System Using CNN Based Traffic Sign Recognition with Integrated CAN Communication

The incidents of deadly traffic accidents in India is rising daily. Recent studies have shown that implementing an efficient speed control system helps to prevent or reduce traffic accidents. An intelligent-based speed regulator system integrated with CAN communication for automobiles is proposed in this study. Image categorization of traffic signs is done using the CNN model. The CNN algorithm's output is transmitted to the speed controller using the CAN communication protocol. Recent research papers have addressed speed governor systems that rely on GPS data. These approaches have several disadvantages in a nation like India where not all regions have access to rapid network coverage. These restrictions can be removed by the suggested model. With 4 convolutional layers and superior performance parameters, the CNN model for classifying traffic signs performs exceptionally well. For the deployed model, the average time taken to process each image is 0.002485 s. The CAN module's latency is in the order of microseconds, and it reduces as the data rate rises. The CNN algorithm performs better than the current models. The governor is able to function effectively and efficiently within a considerable amount of time with the aid of CNN's classifications.