Automatic driving method and system of deep learning based on traffic element visual enhancement

Abstract

The invention discloses an automatic driving method of deep learning based on traffic element visual enhancement. Vehicle driving data is obtained through a vehicle-mounted camera; traffic elements such as vehicles, pedestrians, traffic lights, lane lines, stop lines and the like in the image are identified by utilizing the established perception neural network; visual enhancement is carried out on traffic elements in an image through different color blocks, then the enhanced image serves as input and is imported into a pre-determined and trained deep learning neural network model, and an expected steering wheel angle, accelerator opening and brake force are outputted so as to control a vehicle to achieve automatic driving. The invention further discloses a corresponding system. The attention of the deep learning automatic driving system to the key road traffic elements can be increased, and thus the safety, reliability and robustness of automatic driving are effectively improved.

Description

technical field [0001] The invention belongs to the field of automatic driving of automobiles, and relates to an automatic driving method and system based on deep learning of traffic element visual enhancement. Background technique [0002] The method of using deep learning to realize automatic driving of vehicles is the cutting-edge automatic driving algorithm model in the industry. Generally, the deep learning network is designed first, and then the original image collected by the sensor is used as the input of the deep learning network, and the operations such as braking, acceleration and steering are output through the network as output, and then the deep learning network is trained. Its advantage is that the network model can directly respond to the sensory input, and does not require human intervention to write rules. As long as enough training data is provided, the system can automatically learn driving skills. [0003] However, the existing deep learning automatic d...

AI Technical Summary

Problems solved by technology

[0003] However, the existing deep learning automatic driving methods directly use the original image as the input of the deep learning network, resulting in the network not paying enough attention to traffic elements such as vehicles, pedestrians, traffic lights, lane lines, and stop lines in the image, which may easily cause automatic driving vehicles. Safety accidents, unable to achieve very good automatic driving ability

Images