Area information estimating device, area information estimating method, and air conditioning apparatus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating a person's position in a sensing area from camera images face challenges in accuracy due to variations in image quality, resolution, and physical characteristics, and require complex processing with multiple cameras, increasing costs.
Innovation Solution
A device that uses a camera to detect people by excluding individuals whose image coordinates and sizes do not meet predetermined conditions, focusing on the upper body or whole body detection to calculate accurate positions based on foot coordinates, and employing probability density distributions to reduce detection errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth distance is estimated based on the size of the person in the image, then the existence position of the person can be calculated, but distance estimation accuracy is not stable due to variations in image quality, resolution, physical constitution, and posture
Solution Approach 1:
The invention changes the estimation parameter from using absolute size of the person to using the ratio of upper body size to whole body size. This ratio parameter is more stable across different distances, image qualities, and person characteristics, thereby improving both accuracy and reliability of depth estimation
Solution Approach 2:
The invention introduces an intermediary parameter (the ratio of upper body to whole body size) that mediates between the observable image data and the desired depth information. This intermediary ratio serves as a more reliable indicator than absolute size measurements
2Measurement precision
If two cameras are used to calculate the existence position through triangulation, then the existence position can be determined, but device cost increases and processing becomes complicated
Solution Approach 1:
The invention extracts and utilizes only the necessary information (ratio of upper body to whole body size) from a single camera image, eliminating the need for multiple cameras. This extraction of key discriminatory features allows accurate depth estimation with simplified hardware
Solution Approach 2:
A single camera system is made to perform the function that previously required two cameras. The monocular system achieves depth estimation capability through intelligent analysis of size ratios, making the system universally applicable without requiring stereo camera configurations
Data Source
Figure 1
Figure 2A~2B
Figure 3
AI summary
One aspect of the present invention provides a technology of improving estimation accuracy by a simple method in processing of detecting a person from an image of a camera and estimating information on a sensing area based on a detection result. An area information estimating device includes a camera configured to photograph the sensing area, a person detector configured to detect the person from the image of the camera, and a position calculator configured to calculate an existence position of the person in the sensing area based on a coordinate on the image in which the person is detected by the person detector. At this point, the person detector excludes the person in whom a relationship between the coordinate and a size on the image does not satisfy a predetermined condition from a detection target.