AI detection of garbage disposal on pedestrian islands

Cityscape is a kind of public property. A clean and tidy urban environment is pleasing to the eyes, and can also attract tourists. However, it is found that sometimes when people stop at the red light, they litter the pedestrian islands, cluttering the public environment. The current practice requires inspectors to set up temporary posts to record traffic violations or use mobile cameras for recording, which takes a lot of manpower and time and is costly. Disposal of garbage discarded on pedestrian islands creates a burden on the cleaning personnel. 
As a result, this project is a collaboration between the Department of Environmental Protection and Cloudmatrix. The use of image processing and AI establishes a garbage disposal identification system to capture the images of drivers suspected of or clearly violating the garbage disposal rules, so the Department of Environmental Protection will be able to issue warnings and reports later on.

(Photo provided by the Department of Environmental Protection of Taipei City Government)

Photo: Trial project site (Photos provided by the Department of Environmental Protection of Taipei City Government)

Expected Benefit

The system recognizes and determines the discarded garbage from vehicles, as well as captures and retains images of violations. If the garbage is too small in size with too many varieties, or the surroundings are obstructed by foreign objects, or the disposal motion varies from normal behavior, the system can be used to identify whether the driver really throws waste from the vehicle window. The time of occurrence is provided to the Department of Environmental Protection for further review, which will effectively reduce the manpower and time costs needed for ticketing personnel, and deter people from littering.