Image Processing Apparatus Subject Tracking ID Update

Solution Overview

Problem

Automatic tracking techniques using PTZ cameras often fail when a tracking target crosses another object, resulting in the target being lost and tracking being disrupted.

Innovation Solution

An image processing system that detects subjects in frames, updates the tracking target's identification information if the detected subject's ID changes, allowing the system to continue tracking even after the target is temporarily obscured.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If automatic tracking is performed using a PTZ camera, then tracking of the subject can be automated, but the tracking target is lost when it crosses another object

Engineering Contradiction:

Improveautomatic trackingVSAvoidtracking continuity

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The system performs preliminary actions by detecting subject disappearance in advance and proactively searching for reappeared subjects before tracking is fully lost. When the tracked subject disappears from the detection range, the system doesn't immediately give up tracking but actively searches for the subject when it reappears, maintaining tracking continuity throughout the process.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously monitoring whether the tracked subject appears in detection results and adjusting tracking behavior accordingly. When the subject reappears after disappearing, the system receives feedback about the subject's return and automatically resumes tracking, creating a closed-loop control system that maintains reliable continuous tracking.

Inventive Principle:

Principle #23Feedback

2Ease of operation

If the tracking target crosses another object, then occlusion occurs, but the subject detection fails and tracking is disrupted

Engineering Contradiction:

Improvetracking maintenanceVSAvoidsubject detection

Core Design Contradiction:

Ease of operationVSDifficulty of detecting and measuring

Solution Approach 1:

The system performs preliminary actions by detecting subject disappearance in advance and proactively searching for reappeared subjects before tracking is fully lost. When the tracked subject disappears from the detection range, the system doesn't immediately give up tracking but actively searches for the subject when it reappears, maintaining tracking continuity throughout the process.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously monitoring whether the tracked subject appears in detection results and adjusting tracking behavior accordingly. When the subject reappears after disappearing, the system receives feedback about the subject's return and automatically resumes tracking, creating a closed-loop control system that maintains reliable continuous tracking.

Inventive Principle:

Principle #23Feedback

3Reliability

If the subject reappears after being obscured, then tracking can be restored, but the subject may be assigned new identification information causing tracking loss

Engineering Contradiction:

Improvetracking restorationVSAvoididentification information consistency

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The system implements feedback mechanisms by continuously monitoring whether the tracked subject appears in detection results and adjusting tracking behavior accordingly. When the subject reappears after disappearing, the system receives feedback about the subject's return and automatically resumes tracking, creating a closed-loop control system that maintains reliable continuous tracking.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system applies parameter changes by modifying the subject identification parameters when the tracked subject reappears. Instead of treating the reappeared subject as a completely new object with new ID, the system recognizes it as the same subject and maintains or restores the original identification information, ensuring continuous tracking without ID switching.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20240404076A1Image processing apparatus, image processing method, and non-transitory computer-readable storage medium

Publication Date: 2024.12.05 CANON KK

AI summary

An image processing apparatus comprises one or more memories storing instructions and one or more processors that execute the instructions to detect subjects from a frame, and if no subject having same identification information as identification information of a tracking target subject is detected from a first frame, and identification information of subjects detected from the first frame includes new identification information different from identification information of subjects detected from a second frame, which is a frame before a timing at which the tracking target subject is no longer detected, update the identification information of the tracking target subject to the new identification information.