This paper focuses on the detection of road signs, which is significant for autonomous vehicles such as self-driving cars. Our purpose was to detect traffic signs under various conditions, dealing with issues like color stability, shape-based segmentation, angle rotation, damaged signs, and opacity. We implemented the Residual Network (ResNet) architecture, achieving an accuracy of 97.93%. Experiments show that our ResNet-based detection approach is comparable to standard techniques in terms of real-time performance. The model incorporates machine learning algorithms and techniques such as image preprocessing and image classification. The German Traffic Sign Detection Benchmark (GTSDB) dataset, which contains a wide range of traffic signs under various conditions, was used in our algorithm. The system includes both machine learning techniques for image classification and deep learning algorithms for higher detection rates.