Unmanned sweeper signal lamp fusion perception recognition and control method

Abstract

The invention relates to a signal lamp fusion perception recognition and control method for an unmanned sweeper. According to the method, signal lamps are identified through V2X; obstacles such as pedestrians and vehicles are recognized through the automatic driving sensing component; after analysis of the controller, an instruction is transmitted to the execution unit to control the driving state of the sweeper and the movement of the spraying and sweeping mechanism. According to the invention, the signal lamp is identified through V2X, so that the accuracy is greatly improved, and the false identification condition of a sensing part under an extreme condition is avoided; surrounding obstacles are sensed and recognized through fusion of sensing components such as a camera, millimeter waves and a laser radar, and the false recognition rate is reduced through fusion sensing of multiple sensors; power, steering, braking, and sweeping and spraying actions are performed by fusing the determinations.

Description

technical field [0001] The invention relates to the field of unmanned sweeping vehicles, in particular to a fusion perception recognition and control method for signal lights of unmanned sweeping vehicles. Background technique [0002] With the development of social economy and the advancement of science and technology, unmanned sweeping vehicles have realized commercial operation, which will surely reduce the labor intensity and risks of sanitation workers and greatly facilitate people's lives. However, when an unmanned sweeper encounters a signal light, if it cannot make a timely judgment, it will easily cause traffic safety hazards and reduce traffic efficiency. The current signal light recognition technology mainly uses the camera to identify whether there is a signal light and the color of the signal light. Due to the limitation of the functional principle, it cannot solve the problem of the degradation of its recognition performance in rainy days or ambient light inter...

AI Technical Summary

Problems solved by technology

However, when an unmanned sweeper encounters a signal light, if it cannot make a timely judgment, it will easily cause traffic safety hazards and reduce traffic efficiency.

The current signal light recognition technology mainly uses the camera to identify whether there is a signal light and the color of the signal light. Due to the limitation of the functional principle, it cannot solve the problem of the decline of its recognition performance in rainy days or ambient light interference.

The shortcomings of the existing technology: 1. The imperfection of the algorithm model leads to a significant decrease in the confidence of the camera to identify red and green; 2. The environmental occlusion causes a significant decrease in the confidence of the camera to identify the signal light; 3. The backlight causes a significant decrease in the confidence of the camera to identify the signal light; 4. Insufficient light at night leads to a significant reduction in the confidence of the camera in identifying signal lights; 5. In the case of viaducts or obstacles, whether the GPS signal is available will cause trouble for the driving safety of unmanned vehicles. 6. It is not given Judgment of zebra crossing and stop line

Images