In contemporary times, effective recognition and classification of traffic signs play a crucial role in automation technologies. This article explores a computer vision project implemented using the Python programming language and the TensorFlow library. The project successfully achieves the recognition of traffic signs with an accuracy rate exceeding 95%. TensorFlow is a powerful open-source library that provides capabilities for deep learning model training and recognition. This Python-based project creates a custom neural network model using TensorFlow and optimizes this model through training data. The training process utilizes an extensive dataset of traffic signs, continually refining the model for increased classification accuracy. The obtained results demonstrate that, through effective utilization of TensorFlow, a recognition accuracy exceeding 95% is achieved despite the complexity of traffic sign patterns. This provides a reliable solution for traffic sign recognition applicable in various domains such as driver assistance systems, autonomous vehicles, and traffic safety applications. In conclusion, this study presents a successfully implemented image processing project using the Python programming language and TensorFlow library to achieve high accuracy in the recognition of traffic signs. The results obtained serve as a significant foundation for future research and applications in this field.