A parking lot structure extraction method under the constraint of main direction

Abstract

The invention discloses a method for extracting a parking lot structure under main direction restriction. The method comprises the following steps of: performing initial line segment detection on an aviation orthoimage by using an Edison algorithm and Hough transformation so as to acquire a main direction of a parking space line; detecting a correct parking lot line by using a line segment extraction method under the main direction restriction according to the acquired main direction; grouping and screening the parking lot line according to an angle of the parking lot line and dividing parking lanes by using a maximum intersecting direction method; calculating structural parameters of a parking lot by the extracted parking lot line and the divided parking lanes; and re-building a correct parking space line of the parking lot according to the structural parameters of the parking lot so as to generate a segment line of each parking lane to finish automatic extraction of the parking lot structure. According to the method, the main direction of the parking space line can be acquired by using the initial line segment detection, the parking space line is extracted by using the main direction as the restriction, so that the accuracy, the completeness and the positioning accuracy of the extracted parking space line are high, and an evidence is provided for extracting the parking lot structure; and moreover, the parking lot structure is extracted by using single-scene aviation orthoimages as data, so that the data are easily acquired, and the acquisition cost of the data is low.

Description

technical field [0001] The invention relates to a method for extracting the structure of a parking lot, in particular to a method for extracting the structure of a parking lot under the constraint of the main direction. Background technique [0002] The parking lot, as a place where vehicles temporarily stop, is expanding rapidly with the continuous development of the urbanization process. Today, parking lots are ubiquitous in towns and suburbs alike, even beyond the buildings they serve. As a necessity in the automobile society, parking lots play an irreplaceable role in human production and life, and the cognitive level of parking lots is also particularly important. The knowledge of the structure of the parking lot and the number of parking spaces is directly related to the decision of the minimum parking requirements (MPRs). MPRs have attracted increasing attention in recent years and have become an important topic related to vehicle parking policy. A reasonable vehic...

AI Technical Summary

Problems solved by technology

The main basis for the construction of the parking lot structure is the parking lot line. The correctness and completeness of the parking lot line extraction are directly related to the correctness and accuracy of the parking lot structure extraction. The existing parking lot structure extraction methods are in the parking lot line detection. In the process, the edge detection method is directly used to extract the parking space line. Most of the extracted line segments are not related to the parking space line, and the correctness and integrity are low. Some important parking space line structure information is often missed, which directly affects the construction result of the parking lot structure.

In addition, the original method for data with poor image quality, such as severe lighting changes, obvious shadows, etc., has poor accuracy in extracting the structure of the parking lot

Images