Unmanned vehicle anti-backlight device and method

Abstract

The invention relates to an unmanned vehicle anti-backlight device and method, and belongs to the field of unmanned driving environment perception. The device comprises an image analysis unit, a single chip microcomputer, a motor driving power supply, a motor and a dimming lens, the dimming lens is fixed on the motor; the motor driving power supply is connected with the motor and supplies power tothe motor; the image analysis unit is in signal connection with a single-chip microcomputer and is used for extracting and analyzing an image region of interest and outputting a motor rotation signal; the dimming lens is used for filtering backlight rays; the single-chip microcomputer is connected with the motor and used for converting output signals of the image analysis unit and controlling themotor to drive the dimming lens to rotate. According to the invention, the image quality can be quickly improved in the image acquisition stage, and the device can adapt to backlight scenes with different intensities by adopting the two stages of dimming lenses for filtering.

Description

technical field [0001] The invention belongs to the field of unmanned driving environment perception, and relates to an unmanned vehicle anti-backlight device and method. Background technique [0002] In recent years, with the rapid development of deep learning, visual target detection technology has been widely used in unmanned driving environment perception systems. It uses key feature information such as edges, shadows, and light, or deep feature information of convolutional neural networks to detect targets. People and vehicles can perceive traffic participants such as vehicles, pedestrians, and traffic signs in the surrounding driving environment. However, in sunny mornings and afternoons, strong sunlight seriously interferes with the camera’s metering system, forcing the camera to reduce the exposure, causing the target area on the image to be too dark and the sky area to be overexposed. The picture quality cannot meet the requirements of unmanned visual target detecti...

AI Technical Summary

Problems solved by technology

However, in sunny mornings and afternoons, the strong sunlight seriously interferes with the camera’s metering system, forcing the camera to reduce the exposure, causing the target area on the image to be too dark and the sky area to be overexposed. The picture quality cannot meet the requirements of unmanned visual target detection. Requirements, resulting in the failure of the unmanned driving environment perception system, causing serious traffic accidents

Images