Vehicle Control System Using Dynamic Object Verification for Traffic Light Reliability

The patent introduces dynamic objects (other vehicles and pedestrians) as intermediary elements to verify the lighting state determination. Instead of directly trusting the image-based determination, the system uses the behaviors of these dynamic objects as intermediate verification - if their behaviors are consistent with the determined lighting state, then the determination is considered reliable. This intermediary verification mechanism resolves the contradiction by adding a validation layer without significantly complicating the core determination process.

The system implements feedback by observing whether dynamic objects behave consistently with the determined lighting state. The behavior of dynamic objects serves as feedback information to validate or refute the initial determination. If the feedback indicates inconsistency (e.g., vehicles moving when red light is detected), the system can correct or reject the erroneous determination, thereby improving accuracy while maintaining efficient operation.

Dynamic objects serve as intermediary verification elements that provide indirect validation of the lighting state determination. The system does not directly measure the lighting state with absolute certainty but uses the behaviors of dynamic objects as intermediate evidence. This approach maintains operational simplicity while significantly improving reliability, as the intermediary verification catches erroneous determinations without requiring complex direct measurement improvements.

The system performs preliminary verification by checking whether dynamic object behaviors are consistent with the determined lighting state before finalizing the control decision. This preliminary action using readily observable dynamic object behaviors provides a quick reliability check that maintains fast operation while ensuring safety, avoiding the need for complex verification procedures.

The patent uses dynamic objects as intermediary verification elements that require no additional detection devices. Instead of adding complex verification hardware, the system leverages the naturally occurring dynamic objects in the environment as mediators to validate the image-based determination. This approach maintains low device complexity while improving determination accuracy through intelligent use of existing environmental elements.

The system performs self-verification by using the behaviors of dynamic objects that are already present in the environment. Rather than requiring additional active sensors or verification devices, the system extracts verification information from the self-moving dynamic objects that naturally exist in the traffic scene. This self-service approach improves reliability without increasing device complexity.