Parking space detection method and device, vehicle and storage medium

Abstract

The invention discloses a parking space detection method and device, a vehicle and a storage medium. The method comprises the steps of obtaining a fisheye image; inputting the fisheye image into a pre-trained target neural network to obtain a detection thermodynamic diagram; the detection thermodynamic diagrams comprise a parking space center point thermodynamic diagram, a parking space corner point thermodynamic diagram and a parking space pokeability attribute thermodynamic diagram; and position processing is performed according to the detection thermodynamic diagram to obtain a parking space detection result, so that the fisheye image is input to the target neural network, richer environment information can be obtained, distortion correction is not needed, the image processing flow is greatly simplified, and the detection thermodynamic diagram in the fisheye image is learned by adopting the target neural network, so that the detection efficiency is improved. The parking space detection result is determined through the detection thermodynamic diagram, and compared with the parking space detection result directly determined through the target neural network, the learning task of the target neural network is more visual, a simple structure can be adopted, and the detection speed is improved.

Description

technical field [0001] The present application relates to the technical field of automatic driving, and more particularly, to a parking space detection method, device, vehicle and storage medium. Background technique [0002] With the continuous development of artificial intelligence technology, intelligent vehicles are gradually entering people's daily life. The functions of automatic parking and assisted parking have been applied to some models, and accurate detection of parking spaces is a key step to realize the parking function. [0003] At present, most of the parking space detection methods are based on conventional plane images and neural network models for parking space detection. However, the conventional plane image perception field of vision is limited, and accurate and effective parking space search and identification cannot be performed in complex parking environments. And the parking situation requires the use of a complex neural network model, which is diffic...

AI Technical Summary

Problems solved by technology

[0003] At present, most of the parking space detection methods are based on conventional planar images and neural network models to detect parking spaces. However, conventional planar images have a limited field of view, and cannot perform accurate and effective parking space search and recognition in complex parking environments. And the parking situation requires the use of a complex neural network model, which is difficult to deploy on edge devices and affects the further application of parking technology on vehicles

Images