Method for guiding vehicle driving based on parking space monitoring camera shooting

Abstract

The invention discloses a method for guiding vehicle driving based on parking space monitoring camera shooting. A monitoring camera shooting system, a monitoring image processing system, a digital model of a parking lot scene and a vehicle-mounted intelligent terminal are included. Firstly, accurate digital modeling is carried out on a monitored parking space and a surrounding environment, analysis and modeling are carried out on a vehicle entering a monitoring area, then a real-time position relation between the vehicle and the parking space is analyzed and established from a global perspective, and then auxiliary guidance is carried out on operation of a driver.

Description

technical field [0001] The invention relates to a method for guiding vehicle driving based on monitoring cameras of parking spaces. Firstly, accurate digital modeling is carried out on the monitored parking spaces and the surrounding environment, and the vehicles entering the monitoring area are analyzed and modeled, and then from the overall Angle analysis and establishment of the real-time positional relationship between the vehicle and the parking space, and then assisting and guiding the driver's operation, belong to the technical field of vehicle assisted driving. Background technique [0002] With the increasing popularity of the application of parking space monitoring cameras, it is widely used to realize the status monitoring of parking spaces, unmanned charging and other functions; and, there are also some solutions that use monitoring cameras to monitor the surrounding environment and remind the driver of the vehicle , to prevent accidents from happening. [0003]...

AI Technical Summary

Problems solved by technology

[0005] However, in the prior art, there is no technical solution of "using the parking space monitoring camera to provide precise auxiliary guidance for the driver's operation of entering / exiting the parking space", because there are the following technical problems:

[0006] 1. The problem of camera accuracy: usually the number of pixels of the camera is 2 million pixels and above, which are collectively referred to as high-definition cameras. The number of pixels of the existing high-definition cameras used for monitoring is 3-5 million pixels; Pixel configuration is relatively common; take a 3-megapixel color CCD high-definition camera as an example, the resolution is 2048(H)×1536(V), and the frame rate is 25fps; if the horizontal width of the shooting area is 10 meters, the resolution can be known by calculation It is about 5mm; however, if a larger area needs to be monitored, its resolution will drop. In addition, the camera will inevitably be affected by the shooting angle, and there will be problems of image distortion; therefore, directly monitor various locations through surveillance cameras It is difficult to accurately calculate the relationship and distance, and the amount of data calculation and processing is very large when calculating various positional relationships separately;

[0007] 2. Occlusion problem: No matter how the surveillance camera is installed, or the surveillance cameras installed in multiple locations work together, it is unavoidable that the problem of not being able to capture the overall situation of the vehicle in real time; especially when there are vehicles parked in adjacent parking spaces; The occluded part cannot accurately judge the real-time position relationship through the image, that is, there is a "dead angle" for monitoring; the difference between the roadside camera and the car surround view sensor is that the car surround view sensor is sensed by the car body to the outside, and the probability of being blocked is relatively small; However, no matter how many cameras are installed in the roadside camera system, local occlusion cannot be avoided.

Images