Jiaxun Feihong's Intelligent Video Analysis System for Railways has been certified as a scientific and technological achievement by the Lanzhou Railway Bureau, further boosting the intelligent upgrade of the railway industry!

2025-03-21

Source: Jiaxun Feihong official Weibo

The importance of railway transportation safety has become increasingly prominent with the continuous expansion of railway lines. As an important means of ensuring safety, video surveillance systems have developed rapidly in recent years. By the end of 2024, the national railway integrated video surveillance system had connected more than 260,000 image feeds to the National Railway Group, and these systems are widely used in multiple fields such as transportation command and safety monitoring. However, despite the significant increase in the number of video terminals, the level of intelligence has relatively lagged behind. A vast amount of surveillance data still relies on manual video patrols, making it difficult to improve the efficiency of human resources, and there is an urgent need for technological breakthroughs.

To enhance the efficiency of railway safety management, Beijing Jiaxun Feihong Electric Co., Ltd. (hereinafter referred to as "Jiaxun Feihong," Stock Code: 300213) has joined forces with the Safety Supervision Office of China Railway Lanzhou Group Co., Ltd. (hereinafter referred to as "Lanzhou Railway Bureau") to explore intelligent upgrade pathways and successfully develop the Intelligent Video Analysis System for Railways. As the core solution provider for the project, Jiaxun Feihong has relied on its more than ten years of technical accumulation and experience in the video field. By deeply analyzing the business processes in railway operation control and environmental safety scenarios, the company innovatively established a closed - loop evolution mechanism of "scene - demand - algorithm iteration - value validation." This mechanism has achieved the scenario - based reconstruction of multimodal state - of - the - art (SOTA) algorithms and has tackled the industry - wide challenge of video understanding in complex environments.

The Intelligent Video Analysis System for Railways has upgraded traditional video surveillance with intelligent capabilities, enabling intelligent identification and early warning of safety hazards. This significantly reduces the intensity of manual monitoring and enhances control efficiency. While saving costs, it injects a "smart gene" into railway safety management and ushers in a new chapter of technology - enabled empowerment. Recently, the system passed the technological review of scientific and technological achievements by the Lanzhou Railway Bureau and received high praise from customers, fully demonstrating Jiaxun Feihong's technical strength and innovation capabilities in the field of intelligence.

1903999260469129217.png

The Intelligent Video Analysis System for Railways, independently developed by Jiaxun Feihong, integrates nine major categories and over 20 customized algorithm models, including railway worker cross - line behavior recognition, on - site guard behavior recognition, large - scale mechanical construction operations, and intrusion detection in the throat area. It has conducted targeted research on intelligent monitoring and analysis topics. Since its development began in 2020, the system has undergone nearly four years of innovative exploration and deployment trials, continuously optimizing the models to build a full - stack visual intelligence system that can “see, see clearly, and understand.”1904002032723718145.png

The figure illustrates the application scenarios of the Intelligent Video Analysis System for Railways.

The Intelligent Video Analysis System for Railways boasts robust video analysis and data processing capabilities, functioning like an indefatigable “safety guardian.” It can conduct round - the - clock real - time analysis of abnormal behavior events in a large number of incoming videos, issue timely warnings, and notify relevant personnel in a graded and categorized manner. Through human - machine interactive monitoring, users can confirm and handle the generated warning information, and use on - site loudspeakers to persuade violators to correct their behavior. This reduces the labor intensity of manual remote monitoring and analysis, improves video surveillance efficiency, enhances the intelligence level of video surveillance, and constructs a solid safety barrier for railway construction and maintenance operations. The system can achieve a comprehensive recognition rate of over 95% for scenarios such as personnel违规crossing lines, entering dangerous areas, and违规using mobile phones. It can also conduct intelligent analysis of a large number of historical alarms, automatically generate multi - dimensional alarm trend analysis, and provide data support and scientific basis for safety management decision - making. The transition from “human - to - human monitoring” to “AI - based monitoring” helps further upgrade the intelligence of railways.

At present, the Intelligent Video Analysis System for Railways has been deployed and tested at 28 stations on the Tianshui - Lanzhou section of the Longhai Line, triggering over 50,000 effective alarms. This has significantly reduced the disruption of abnormal events to railway transportation order and effectively lowered railway transportation safety risks. It can be said that this system is not only a “protector” of railway safety but also an “intelligent pioneer” in discovering and preventing potential hazards.

The rapid development of artificial intelligence technology is reshaping traditional industries, and its potential to improve efficiency and optimize processes has been widely verified. However, in complex scenarios of vertical industries such as railways, algorithm development still faces many challenges. Based on a deep understanding of the railway industry and years of accumulated scene - based data, Jiaxun Feihong has deeply integrated artificial intelligence with big data technology. Through customized algorithm models and innovative solutions, it has not only broken through the technical bottlenecks in complex scenarios but also provided strong support for the intelligent upgrade and high - quality development of railways.

Jiaxun Feihong has always been committed to the mission of “making industries safer, more efficient, and more intelligent.” It continuously delves into railway industry scenarios and creates value for customers with dedication. By actively responding to and providing customized technical support and solutions, it has won the trust and support of many customers. In the video “intelligent monitoring” project on the Tianshui - Lanzhou section of the Longhai Line of the Lanzhou Railway Bureau, Jiaxun Feihong also received a letter of commendation from the Lanzhou Railway Bureau for its outstanding technical strength and spirit of service. The letter highly praised the project team members for their proactive, dedicated, and patient - and - responsible attitudes at work.

The successful passing of the technological review of scientific and technological achievements by the Lanzhou Railway Bureau for the Intelligent Video Analysis System for Railways fully reflects the bureau's forward - looking layout and firm determination in promoting the intelligent transformation of railways. It also highlights Jiaxun Feihong's mission and responsibility to empower industry development with technological innovation. The close collaboration between the two parties has opened a new chapter in technology - enabled railway safety and provided a replicable and promotable innovative model for the high - quality development of the railway industry.

In the future, Jiaxun Feihong will continue to be driven by technological innovation and join hands with more industry customers to jointly promote the shift of railway safety management from “passive response” to “active prevention.” It will accelerate digital transformation and contribute more to building a strong transportation nation and achieving sustainable industry development.