Traffic Big Data Intelligent Object Tracking Technology under Noise Conditions

Intelligent transportation systems make extensive use of image processing techniques such as object detection and tracking. Real-world traffic conditions require specialized processing of image noise, typically caused by weather conditions or occlusions. This paper proposes a contextual community attention graph neural network structure, which uses the attention mechanism representation of different community nodes to associate important recognition features. This structure can effectively track objects in noisy videos. Algorithm comparison experiments on the dataset show that the method performs better in rainy and night conditions.