<![CDATA[Tires are a crucial component of vehicles that play an important role. The functions of tires include reducing vibrations from road irregularities, protecting the wheels from wearing out quickly, and providing ease of movement while driving. Due to their vital nature, it is important to maintain the condition of tires to ensure passenger safety and comfort. Excessive tire usage can lead to damage such as cracks. Cracks in tires can occur due to poor weather conditions and road conditions. Cracks in tires refer to a condition where the tire loses its flexibility and traction capabilities while driving. In fact, research data shows that 80% of traffic accidents on highways occur due to indications of tire damage. Prompt handling and regular checks are required to address and optimize tire damage. The methods used to check tire conditions previously were done manually and relied on human labor. These methods are considered ineffective in identifying tire cracks. In this study, a Deep Learning model using the Transfer Learning ShuffleNet approach was developed to automatically classify tire images in identifying tire cracks. The main objective of this research is to determine the best method in identifying tire cracks and measure the performance of the developed model. In the development of this model, testing was conducted using 10 different scenarios on the created model to find the best method for achieving optimal testing accuracy. The best results obtained were an accuracy of 78% using the ADAM optimizer and 75% using the RMSprop optimizer. Therefore, it can be concluded that the Transfer Learning ShuffleNet method is efficient and capable of accurately detecting tire cracks. This research also successfully determined the best parameters such as the number of epochs, dropout layers, and optimizer in model creation to achieve optimal results. Through the adoption of Transfer Learning ShuffleNet, this research contributes to the development of tire damage detection technology aimed at improving safety and driving comfort.]]>
Copyrights © 2024
