Multimodal Object Recognition for Non-Cooperative Personnel Tracking

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing personnel recognition technologies are inefficient and inaccurate in densely populated areas, causing congestion and affecting traffic efficiency, especially during peak periods, and require cooperative and perceptual methods.

Innovation Solution

An object recognition system integrating face recognition, gait recognition, pedestrian re-recognition, WiFi probe recognition, and iris recognition technologies to accurately identify individuals in complex scenes without cooperative interaction, using modules for information collection, video acquisition, and recognition.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional close and cooperative personnel information collection methods are used, then personnel identity recognition can be achieved, but traffic efficiency is reduced and congestion occurs during peak periods

Engineering Contradiction:
Improvepersonnel identity recognition accuracyVSAvoidtraffic efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces traditional mechanical close-range cooperative recognition methods with remote non-cooperative recognition technology. The system uses cameras, sensors, and computer vision algorithms to automatically capture and analyze personnel information from a distance, eliminating the need for physical interaction or close proximity, thus maintaining recognition accuracy while improving traffic flow efficiency

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

Solution Approach 2:

The patent introduces an intermediary recognition system that acts as a mediator between personnel and the verification target. Instead of direct close-range interaction, the system uses intermediate devices such as cameras, image processing units, and data transmission systems to remotely capture, process, and verify personnel identity information, thereby resolving the contradiction between recognition accuracy and traffic efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If multiple recognition methods are integrated to improve recognition accuracy in complex scenes, then personnel identification becomes more accurate, but system complexity increases

Engineering Contradiction:
Improvepersonnel identification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple recognition methods including face recognition, gait recognition, iris recognition, and WiFi probe recognition into a unified integrated system. By combining these different recognition technologies, the system achieves high accuracy in complex scenes while managing overall system complexity through unified architecture and coordinated operation of various modules

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent creates a universal recognition system that can perform multiple recognition functions (face, gait, iris, WiFi) within a single platform. This multi-functional system handles diverse recognition tasks using a common framework, reducing the need for separate systems for each recognition method and thereby controlling system complexity while maintaining high identification accuracy

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

Data Source

PatentEP3796214B1Object recognition system and method
Publication Date: 2025.08.13 NUCTECH CO LTD
  • EP3796214B1 patent drawingFigure 1~2
  • EP3796214B1 patent drawingFigure 3~4
  • EP3796214B1 patent drawingFigure 5

AI summary

The present disclosure provides an object recognition system and a method thereof, the object recognition system can include: an information collection module configured to acquire information about the object to be inspected; a first video acquisition module configured to acquire an image of the object to be inspected, and acquire multiple types of features of the object to be inspected by using a plurality of recognition methods based on the image of the object to be inspected; a storage module configured to store the information about the object to be inspected in association with the multiple types of features of the object to be inspected; a second video acquisition module configured to track and capture a suspect object to acquire an image of the suspect object, and acquire multiple types of features of the suspect object by using the plurality of recognition methods based on the image of the suspect object; and a recognition module configured to recognize identity information of the suspect object based on the multiple types of features of the suspect object and the multiple types of features of the object to be inspected which are stored in association with the information about the object to be inspected in the storage module.