Home researchers have developed synthetic intelligence (AI)-based CCTV know-how for implementing sensible cities in Vietnam’s transportation sector. The developed know-how is scheduled to be pilot-tested in some Vietnamese cities and utilized for understanding site visitors scenario standing.
The Korea Institute of Civil Engineering and Constructing Expertise (KICT) introduced on Nov. 11 that it has developed AI-based clever CCTV know-how that displays street site visitors traits for sensible metropolis localization in Vietnam’s transportation sector.
Within the transportation subject, as demand for superior applied sciences similar to AI and large data-based real-time monitoring and incident detection will increase, the need for clever CCTV able to real-time detection of threat components similar to wrong-way autos and site visitors accidents is rising. Such know-how is centered on AI deep learning-based algorithms, making large-scale, high-quality studying information important. The Vietnamese authorities is conducting pilot tasks in main cities together with Hanoi, Da Nang, and Ho Chi Minh Metropolis to regularly develop sensible metropolis implementation by 2030, however domestically, there have been limitations in that information acquisition shouldn’t be simple and information assortment consumes appreciable time and prices.
Accordingly, the KICT ITS Efficiency Analysis Heart analysis staff utilized switch studying strategies to attain excessive efficiency even in restricted native studying information environments, presenting sensible options for sensible metropolis localization. To complement inadequate studying video information, the analysis staff utilized high-quality information secured domestically and retrained algorithms optimized for Vietnamese native site visitors circumstances. Primarily based on this, they respectively developed Vietnamese site visitors object recognition algorithms, site visitors scenario judgment algorithms, and incident detection algorithms. The developed algorithms had been put in in clever CCTV prototypes and deployed at three sensible metropolis venture websites in Tam Ky Metropolis, Quang Nam Province, Vietnam, via which subject testing operations and efficiency verification had been carried out.
The developed clever CCTV demonstrated excessive accuracy of 99.6% in object recognition of pedestrians, bikes, and cars. Incident detection accuracy additionally reached 97.8%, proving excessive efficiency. By way of this know-how growth, KICT has secured core transportation infrastructure know-how that aligns with Vietnam’s sensible metropolis technique. That is anticipated to contribute to strengthening the technological competitiveness of home small and medium enterprises and establishing a basis for abroad market entry.
In the meantime, this achievement was developed via the venture “Improvement and Native Software of Vietnam Sensible Metropolis-Specialised Clever CCTV Expertise,” a KICT small and medium enterprise assist venture funded by the Ministry of Science and ICT, and the related know-how was transferred to Yoongchang Co., Ltd. Director Park Solar-kyu said, “This analysis achievement considerably contributes to the sensible localization of Vietnam’s sensible metropolis transportation sector,” and “KICT will proceed to actively assist home corporations’ abroad enlargement and lead in international sensible metropolis know-how.”
