Object Detection Using Geometrical Coefficients for Lighting Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing outdoor object detection systems for controlling sidewalk lighting at night suffer from high detection error rates, limited detection range, and dependence on operating conditions, with most systems unable to differentiate between types of moving objects.
Innovation Solution
A method for robust object detection in a sequence of images, particularly in dark street environments, using geometrical parameters analysis, including area, orientation, and extent coefficients, to identify pedestrians and cyclists, and adjust lighting accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If histograms of oriented gradients are used for object detection, then detection accuracy is improved, but computational complexity increases making it inapplicable for low-performance SoCs
Solution Approach 1:
The patent extracts only the essential geometrical features (bounding box coordinates, area, orientation, extent coefficients) needed for detection, discarding the computationally intensive histogram calculations while maintaining detection accuracy for pedestrian and cyclist identification
Solution Approach 2:
The patent replaces complex computational algorithms with simple geometrical calculations that can be performed efficiently on low-performance hardware, using basic mathematical operations instead of intensive machine learning models
2Device complexity
If frame differencing or background subtraction is used for object detection, then computational power requirements are reduced, but the method cannot differentiate between different types of objects and has high false-positive error rate
Solution Approach 1:
The patent applies different analysis methods to different aspects of the detected objects: background subtraction is used for motion detection while geometrical parameter analysis (area coefficient, extent coefficient, orientation) is applied specifically to differentiate between pedestrian and cyclist types, combining the advantages of both approaches
Solution Approach 2:
The patent changes the parameters used for object characterization from simple presence/absence detection to detailed geometrical parameters including bounding box dimensions, area coefficients, and orientation angles, enabling differentiation between object types while maintaining low computational requirements
3Ease of manufacture
If passive infrared sensors are used for motion detection, then device cost and power consumption are reduced, but detection capability is limited by ambient temperature and brightness conditions
Solution Approach 1:
The patent introduces an image sensor as an intermediary that captures visual information and processes it through background subtraction and geometrical analysis, mediating between the simple motion detection capability and the need for reliable object identification under varying environmental conditions
4Measurement precision
If radio wave sensors are used for object detection, then detection sensitivity is improved, but false detections increase due to sensitivity to ambient conditions
Solution Approach 1:
The patent implements dynamic adjustment of detection parameters and background models that adapt to changing environmental conditions, allowing the system to maintain high sensitivity while reducing false positives by learning and adapting to ambient variations over time
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The proposal concerns a method, an apparatus, and a computer program comprising program code for detecting a wanted object in a sequence of images captured by an image sensor. Information about a bounding rectangle of a moving object in the sequence of images and information about a contour area of the moving object are obtained (10). Geometrical coefficients are then determined (11) for the moving object using the information about the bounding rectangle and the information about the contour area. Based on at least the geometrical coefficients it is determined (12) whether the moving object is a wanted moving object.