| Sumari: | Self driving cars are an interesting topic to discuss due to the high level of traffic accidents that occur due to human error. Self driving cars are vehicles that can find out about the environment with minimal human intervention. Self driving itself has many development methods such as Light Detection and Ranging (LIDAR), cameras, radars, or a combination of these sensors. This study made a prototype self-driving car using a camera as a sensor and a neural network algorithm for pattern recognition. The pattern recognition in question is the image recognition of the path taken. The data that has been taken will later be converted into a matrix with dimensions of 320x120 according to the image resolution. Then the data will be trained to recognize the path pattern with the proportion of 7: 3 for training accuracy and validation accuracy. The resulting prediction has an accuracy of 76.86% for training accuracy and 75.24% for validation accuracy.
|
|---|