Frame-Matched Image Inspection for Accurate Object Counting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image sensors face challenges in accurately counting objects passing through a capturing field of view due to redundant or omitted counts, especially when multiple objects are aligned or close together, leading to inefficiencies in programming and verification processing.

Innovation Solution

An image inspection device that continuously captures a field of view to generate multiple frame images, performs inspection processing, and sets a learning image window to detect target regions, matching them across frames, and counts objects based on positional information and overall moving direction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the frame rate is increased to prevent capturing omission, then the reliability of object counting is improved, but the device complexity and processing load increase

Engineering Contradiction:
Improveobject counting accuracyVSAvoidprocessing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the object detection task by identifying and extracting only the necessary frame images (first frame image and second frame image) from the time-series sequence, rather than processing all frames. This selective processing reduces computational load while maintaining counting accuracy through the matching process between these key frames.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by detecting target regions in advance and storing their positional information before the actual counting operation. The inspection execution section detects target regions and stores positional information in advance, enabling efficient matching and counting without requiring complex real-time processing of all frames.

Inventive Principle:
Principle #10Preliminary action

2Ease of operation

If simple ON/OFF counting is used, then the ease of operation is improved, but the measurement precision deteriorates due to redundant or omitted counts

Engineering Contradiction:
Improvecounting operation simplicityVSAvoidobject counting accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent implements feedback through the matching process that compares positional information between the first frame image and second frame image. The inspection execution section matches target regions between frames based on positional information and overall moving direction, providing feedback to correct and verify counting results, thereby eliminating redundant or omitted counts while maintaining operational simplicity.

Inventive Principle:
Principle #23Feedback

3Reliability

If multiple frame images are processed, then the reliability of object detection is improved, but the loss of time increases due to processing multiple frames

Engineering Contradiction:
Improveobject detection accuracyVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts only the essential frame images (first frame image and second frame image) from the time-series sequence of captured images. By taking out only these critical frames for processing rather than handling all frames, the system maintains detection reliability through the matching process while significantly reducing processing time and computational resources.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS20250299317A1Image inspection device
Publication Date: 2025.09.25 KEYENCE CORP
  • US20250299317A1 patent drawing
  • US20250299317A1 patent drawing
  • US20250299317A1 patent drawing

AI summary

An image inspection device includes an image capturing section that generate a plurality of frame images aligned in time series, an inspection execution section that outputs an inspection result, and an inspection setting section that performs setting of the inspection execution section. The inspection setting section receives setting of a window for a learning image on which the object appears The inspection execution section detects target regions from a first frame image and a second image based on the learning image and the window set for the learning image, performs matching processing of the target region between the first frame image and the second frame image based on the detected target regions from the first image frame and the second image frame, and an overall moving direction of the object, and counts the object based on a result of the matching processing.