Camera Position Evaluation Using Resolution and Gazing Point Metrics

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing camera position evaluation devices cannot effectively assess the suitability of a camera's position and posture for image processing, particularly in terms of object detection and identification, as they rely on inappropriate evaluation metrics such as the area of a triangle formed by the viewer, camera, and point of view, which do not accurately represent the camera's state from the perspective of image processing suitability.

Innovation Solution

A camera position posture evaluating device that calculates a comprehensive evaluation index based on resolution and gazing point angle suitability rates, using resolution and gazing point angle evaluation functions to determine how well an object appears in an image for various camera positions and postures, incorporating penalty coefficients for partial object protrusion or obstruction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the area of a triangle formed by viewer position, camera position, and point of view is used as an evaluation metric, then the similarity between camera field of view and viewer field of view is improved, but the suitability for image processing (object detection and identification) deteriorates

Engineering Contradiction:
Improvefield of view similarity measurementVSAvoidimage processing suitability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent changes the evaluation parameters from geometric metrics (triangle area, angles) to image processing-oriented metrics (resolution suitability rate, gazing point angle suitability rate). This involves transforming the evaluation criteria to match the actual requirements of object detection and identification tasks, thereby resolving the contradiction between field of view similarity and image processing suitability.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If general camera position evaluation methods are used, then the evaluation process is simple, but the evaluation accuracy for image processing suitability deteriorates

Engineering Contradiction:
Improveevaluation process complexityVSAvoidcamera state evaluation accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the evaluation process into distinct components: resolution evaluation (assessing whether the target object fits within the image frame) and gazing point angle evaluation (assessing the camera's viewing angle relative to the target). This segmentation allows for a more accurate and comprehensive evaluation of camera position suitability for image processing while maintaining a structured and manageable evaluation framework.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2802149B1Camera position/posture evaluation device, camera position/posture evaluation method, and camera position/posture evaluation program
Publication Date: 2020.03.18 NEC CORP
  • EP2802149B1 patent drawingFigure 1
  • EP2802149B1 patent drawingFigure 2~3
  • EP2802149B1 patent drawingFigure 4

AI summary

There is provided a camera position posture evaluating device which can calculate a value which indicates an evaluation of a state of a camera from a viewpoint of to what degree an object appears in an image suitably for image processing. A resolution evaluating means 71 calculates a first suitability rate matching each position in a target area from a viewpoint of resolution. A gazing point angle evaluating means 72 calculates a second suitability rate matching each position in the target area. A target area evaluating means 73 calculates a third suitability rate from a viewpoint of how a target object field-of-view area and the target area overlap. A comprehensive evaluating means 74 calculates an evaluation index which indicates to what degree the state of the camera is suitable to the image processing based on the first suitability rate matching each position in the target area, the second suitability rate matching each position in the target area and the third suitability rate.