Pedestrian Detection Using Top-View Height Verification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing pedestrian detection systems using a single normal camera face high costs and produce many false positives, while solutions involving multiple cameras or other sensors are expensive and complex.

Innovation Solution

A pedestrian detection system that uses a fish-eye lens camera to obtain images, applies distortion correction, transforms the image into a top view, calculates the actual height of pedestrian candidates using extrinsic camera parameters, and generates alerts for true positives within a predetermined height range.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple cameras or other sensors are used for pedestrian detection, then detection reliability is improved, but device complexity and cost increase

Engineering Contradiction:
Improvepedestrian detection reliabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent transforms the 2D image coordinates into a top-view 3D coordinate system by calculating real-world distances and heights. This dimensional transformation enables accurate pedestrian detection using only a single 2D camera, achieving reliability comparable to multi-sensor systems without the added complexity

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent creates a virtual top-view copy of the real scene by mathematically transforming image coordinates into a top-view coordinate system. This virtual copy allows the system to reason about pedestrian positions and heights as if viewed from above, enabling accurate detection with a single camera

Inventive Principle:
Principle #26Copying

2Device complexity

If conventional single camera solutions are used for pedestrian detection, then cost is reduced, but false positive rate increases

Engineering Contradiction:
Improvesystem costVSAvoidfalse positive rate
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent replaces complex multi-sensor mechanical systems with a computational approach using a single camera. By substituting physical multiple sensors with mathematical transformations and calculations, the system achieves high detection accuracy while maintaining low cost and simple hardware configuration

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the parameter space by calculating real-world physical parameters (distance, height, position) from 2D image coordinates. This parameter transformation enables the system to distinguish true pedestrians from false positives by verifying their physical dimensions match expected ranges

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If complex processing algorithms are used to reduce false positives, then detection precision is improved, but computation complexity increases

Engineering Contradiction:
Improvepedestrian detection precisionVSAvoidcomputation complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential geometric relationships needed for pedestrian detection from complex image processing. By focusing on calculating distance and height from top-view coordinates, the system achieves high precision while avoiding unnecessary computational complexity from elaborate algorithms

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2926317B1System and method for detecting pedestrians using a single normal camera
Publication Date: 2020.02.12 HARMAN INT IND INC
  • EP2926317B1 patent drawingFigure 1
  • EP2926317B1 patent drawingFigure 2A~2B
  • EP2926317B1 patent drawingFigure 3A~3B

AI summary

The present application provides pedestrian detection system and method. A pedestrian detection method includes: obtaining an image captured by a camera; identifying a pedestrian candidate in the image; transforming the image into a top view image; calculating the actual height of the pedestrian candidate based on the top view image and extrinsic parameters of the camera; and determining whether the pedestrian candidate is a true positive by determining whether the actual height of the pedestrian candidate is within a predetermined pedestrian height range. The system and method of the present application have lower cost and higher accuracy compared with conventional technologies.