Multi-Camera Occupant Detection Using Timing Verification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for detecting the number of persons on a vehicle are limited in accuracy and efficiency, particularly when occupants are partially hidden or imaged from a single direction, leading to incomplete detection and increased processing loads.

Innovation Solution

An analysis apparatus utilizing multiple cameras capturing images from different directions to detect human faces and determine the number of occupants, with a determination unit verifying if images satisfy a predetermined condition based on imaging time, thereby improving detection accuracy and reducing processing loads by integrating data from multiple angles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple cameras capture images from different directions, then detection accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvedetection accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system divides the detection task into separate modules: a first camera captures images from a first direction, a second camera captures images from a second direction, and the determination unit processes each image group independently before integrating results. This segmentation allows multiple cameras to be used without overwhelming complexity in a single processing unit.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a temporal dimension by associating image groups with timing information (first timing for first camera, second timing for second camera). The determination unit uses these timing parameters to verify whether images satisfy predetermined conditions, adding a time-based filtering dimension that reduces complexity by pre-filtering irrelevant images before integration.

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

2Productivity

If images are processed without timing verification, then processing speed is maintained, but unnecessary processing increases computational load

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidcomputational processing
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The determination unit performs preliminary verification by checking whether the first and second image groups satisfy predetermined conditions based on timing information before proceeding to detect the number of occupants. This preliminary action filters out irrelevant image pairs, reducing the computational load on subsequent processing steps.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses timing information as feedback to verify whether images should be processed together. The determination unit compares the first timing and second timing to determine if the images satisfy predetermined conditions, creating a feedback mechanism that guides the processing workflow and avoids unnecessary computations on mismatched images.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If occupants are detected from a single direction, then device complexity is reduced, but detection accuracy decreases due to hidden occupants

Engineering Contradiction:
Improvedetection accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges the detection results from multiple cameras by integrating the number of occupants detected from the first direction with the number of occupants detected from the second direction. The determination unit combines these separate detection results to arrive at a comprehensive count, ensuring all occupants are counted even if hidden in one view.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The determination unit serves multiple functions: it verifies timing conditions, determines whether image groups satisfy predetermined conditions, and integrates detection results from multiple cameras. This multi-functionality allows the system to achieve high detection accuracy without proportionally increasing device complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3654276B1Analysis device, analysis method, and program
Publication Date: 2023.09.13 NEC CORP
  • EP3654276B1 patent drawingFigure 1
  • EP3654276B1 patent drawingFigure 2
  • EP3654276B1 patent drawingFigure 3

AI summary

The present invention provides an analysis apparatus (10) including a first image analysis unit (11) which detects persons from a first image generated by a first camera that captures a vehicle passing through a road; a first detection unit (12) which detects the number of occupants in the vehicle captured in the first image; a second image analysis unit (13) which detects persons from a second image generated by a second camera that captures the vehicle passing through the road; a second detection unit (14) which detects the number of occupants in the vehicle captured in the second image; a determination unit (15) which determines whether or not the first image and the second image satisfy a predetermined condition; and a decision unit (16) which decides, in a case where the first image and the second image satisfy the predetermined condition, the number of occupants in the vehicle with use of the number of occupants in the vehicle captured in the first image and the number of occupants in the vehicle captured in the second image.