According to media reports, Chengdu programmer Pu Haiyang recently released a vehicle violation reporting program that integrates AI visual recognition technology, aiming to improve traffic supervision efficiency through technological means. The project was conceived in 2025, and Pu Haiyang completed the demo version in just one week during this year's Spring Festival holiday. He used AI tools to generate a web version in three days. Currently, the development of the Android and iOS client versions has reached 80%, and it is expected to officially launch on app stores within two to three months.

The program's core technical advantage lies in compressing the traditional long-cycle manual reporting process into just a few seconds. The system integrates high-precision AI models, supporting automatic identification of lane lines and traffic lights, and accurately distinguishing between motor vehicle and non-motor vehicle lanes, thus enabling automatic classification and determination of violations.

Notably, Pu Haiyang introduced an "auto-zoom and target tracking" function in the App, ensuring that the camera can lock onto violators and maintain high-precision recognition performance even in complex weather or lighting conditions. The overall recognition accuracy has now exceeded 90%.

In terms of data security and compliance, the program adopts a local storage solution, where the original video recorded by users does not go through the developer's server but directly connects to the official interface of the traffic management department. At the same time, the system retains a secondary manual confirmation step, using AI pre-screening combined with manual final review to avoid risks of algorithmic misjudgment.

Although the project has sparked discussions on the boundaries of "public supervision" on social platforms, from an industry perspective, it marks the deep penetration of edge-side AI visual technology in vertical fields of people's livelihood. If such applications can successfully connect with official data systems in the future, they will effectively fill the gaps in public supervision and drive the transformation of urban traffic governance toward digitalization and public participation.