Parking Lot Image Processing Apparatus for Accurate Vehicle Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vehicle detection methods using image processing in parking lots often experience determination errors due to the influence of adjacent parking regions, leading to incorrect identification of vacant or occupied states.
Innovation Solution
An image processing apparatus that acquires images of parking lots, sets determination target regions for each parking space to exclude areas shielded by vehicles in adjacent spaces, and uses edge detection to determine the presence or absence of vehicles within these regions, thereby isolating the parking state of each space from adjacent influences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire parking region is used as the determination target region, then the detection coverage is maximized, but determination errors occur due to noise from adjacent parking regions
Solution Approach 1:
The parking region is divided into multiple determination target regions (first, second, third, and fourth determination target regions) based on the relative positions of adjacent parking regions. This segmentation allows each region to be analyzed independently, preventing noise from adjacent regions from affecting the detection accuracy in each specific area.
Solution Approach 2:
Different determination target regions are defined with different spatial characteristics based on their local environment. The first determination target region excludes the shielded region from the second parking region, while the second determination target region excludes the shielded region from the first parking region. This local adaptation ensures optimal detection accuracy for each specific location.
2Measurement precision
If vehicle detection devices are installed in each parking region, then detection accuracy is improved, but the device quantity and system complexity become enormous
Solution Approach 1:
A single image capturing device performs the function of multiple detection devices by capturing images of multiple adjacent parking regions simultaneously. The image processing apparatus then analyzes these images to determine vehicle presence in each parking region, making the single device multi-functional and eliminating the need for numerous individual detection devices.
Solution Approach 2:
The image processing apparatus segments the captured image into multiple determination target regions corresponding to different parking regions. By processing each segment independently with appropriate noise exclusion, the system achieves multi-region detection accuracy equivalent to having multiple devices while using only one capturing device.
3Area of stationary object
If adjacent parking regions are captured simultaneously, then the overall parking lot coverage is improved, but determination errors occur due to mutual interference
Solution Approach 1:
The captured image is divided into multiple determination target regions, each corresponding to a specific parking region. This segmentation allows simultaneous coverage of multiple parking regions while maintaining independent analysis for each region, preventing mutual interference from causing determination errors.
Solution Approach 2:
Shielded regions that cause interference are extracted and excluded from the determination target regions. By removing these problematic areas from the analysis, the system maintains high reliability in determining vehicle presence while still capturing and analyzing multiple adjacent parking regions simultaneously.
Data Source
AI summary
An image processing apparatus includes a computer, a memory, and a program stored in the memory that, when executed by the computer, causes the computer to perform acquiring an image obtained by capturing a parking lot including a plurality of parking regions capable of parking a plurality of automobiles, deciding, for the image, a determination target region for each of the plurality of parking regions so as to exclude a shielded region that is shielded by an automobile in a case in which the automobile is parked in another parking region, and detecting, from the image, an automobile in each determination target region, and determining, based on a result of the detection, the presence/absence of parking of an automobile in each of the plurality of parking regions.