System and method for determining obstacle by using aerial view images

Abstract

The invention provides a system and a method for determining an obstacle by using aerial view images, relates to technology for determining the obstacle around a vehicle by aerial view image technology. A plurality of image acquisition devices which are arranged at different positions are used for acquiring a plurality of images around a vehicle body respectively, and the plurality of images are at least provided with an overlapping region; an image processor performs independent aerial view image conversion on the images; and an obstacle calculating unit compares the correlation of repeat regions of two independent aerial view images and judges whether the obstacle exists or not according to the size of the correlation. In addition, a mapping table of a group of space conversion data is constructed and comprises driving state data such as the steering wheel angle, gear or speed of the vehicle, and the like, so when the driving state is changed, a panoramic aerial view image at an appropriate viewing angle is quickly converted through the mapping table, and the position of the obstacle measured by the obstacle calculating unit is marked on the panoramic aerial view image; therefore, the safe driving of a user is facilitated.

Description

technical field [0001] The present invention relates to an obstacle detection system, in particular to an obstacle detection system used when driving a vehicle. Background technique [0002] The cost of cameras is decreasing day by day, and photographic equipment is now widely used in vehicle driving assistance, but most of them only passively provide images around the vehicle to assist drivers in avoiding accidents. Since it is difficult for the driver to monitor multiple photographic images at the same time, and the visual blind spot is still large when the traditional planar image is used for parking assistance, there are currently some companies in the industry that can use multiple cameras placed around the car body to combine the multiple images obtained. The images are integrated and converted to form a bird's-eye view image system of a virtual surround field viewed from the roof, which is more in line with the visual habits of the human eye and can further overcome t...

AI Technical Summary

Problems solved by technology

As disclosed in the patent application with the patent publication number JP2007041791 (A), a system and method for displaying a bird's-eye view image, which first uses multiple cameras to collect images around the vehicle, stitches the images into a bird's-eye view image of the surrounding field, and learns from the images Whether there are obstacles near the front, rear, left, and right of the vehicle, and in order to overcome the problem that it is difficult to display high-level obstacles at the image stitching, the patent uses four ultrasonic devices to detect the dead-angle area to solve the problem

However, although this method can overcome the above problems, it will relatively increase the cost of the ultrasonic device

[0005] Furthermore, in the conventional two-dimensional bird's-eye view surround field imaging system, the virtual viewpoint is mostly set above the vehicle. However, when the vehicle is moving forward, backward or turning, the size of the field of view of the image area in the direction the driver needs to move is also different. , but the conventional technology only uses a bird's-eye view image with a fixed virtual perspective, which is difficult to meet the driver's needs

Images