Identity Recognition via Target Feature Mediation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing methods for human identity recognition in video surveillance face accuracy issues when the facial region is unclear or obscured, leading to inaccurate identity determination.

Innovation Solution

A method that extracts a target feature from an image and searches for a corresponding facial information item based on a pre-established correspondence between target features and facial information items, allowing for identity determination without the need to extract facial features from the image, even if the facial region is unclear or obscured.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If facial feature extraction is used for identity recognition, then the method is simple and direct, but the recognition accuracy deteriorates when the facial region is unclear or obscured

Engineering Contradiction:
Improvesimplicity of recognition methodVSAvoididentity recognition accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent introduces target features (such as clothing, accessories, or body characteristics) as an intermediary element between the image and identity recognition. Instead of directly relying on facial features, the system extracts target features from the image, searches for corresponding facial information items using these target features as mediators, and then determines identity. This intermediary approach allows recognition to proceed even when facial regions are obscured or unclear.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent performs preliminary extraction of target features from the image before attempting facial feature extraction. By first identifying target features (such as clothing patterns, accessories, or distinctive body characteristics), the system prepares alternative recognition pathways in advance. This preliminary action ensures that if facial features are unavailable or unclear, the system already has target feature data to facilitate identity recognition through other means.

Inventive Principle:
Principle #10Preliminary action

2Ease of operation

If facial feature extraction is required, then the recognition process is straightforward, but the system becomes vulnerable to obscured or unclear facial regions

Engineering Contradiction:
Improvestraightforwardness of recognition processVSAvoidrobustness against obscured facial regions
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent changes the recognition parameter from facial features alone to a combination of target features and facial information items. By shifting the primary extraction parameter from facial region characteristics to target object characteristics (such as clothing, accessories, or body features), the system maintains operational straightforwardness while improving reliability. The search module uses these changed parameters (target features) to query the database, enabling robust recognition even when traditional facial parameters are compromised.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If direct facial feature matching is used, then the processing is efficient, but the identification accuracy deteriorates when facial regions are not well-defined

Engineering Contradiction:
Improveprocessing efficiencyVSAvoididentification accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the identity recognition process into two independent stages: target feature extraction and facial information item search. Instead of performing direct facial feature matching in a single step, the system first extracts target features (such as clothing characteristics or accessories) and then uses these segmented features to search for corresponding facial information items in the database. This segmentation allows efficient processing of target features while improving identification accuracy by providing an alternative pathway when facial regions are not well-defined.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3678047B1Method and device for recognizing identity of human target
Publication Date: 2023.07.19 HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
  • EP3678047B1 patent drawingFigure 1~2
  • EP3678047B1 patent drawingFigure 3~4

AI summary

The embodiments of the present application provide a method and apparatus for recognizing an identity of a human target. The method includes: obtaining a to-be-recognized image; extracting a target feature of a to-be-recognized human target in the to-be-recognized image as a to-be-searched target feature; searching for a facial information item corresponding to the to-be-searched target feature based on a pre-established correspondence between target features and facial information items; and determining an identity of the to-be-recognized human target based on the found facial information item. It can be seen that, in this solution, there is no need to extract the facial features in the image, and even if the facial region is not clear or is shaded by other objects in the image, the accuracy of identity recognition is reduced. Therefore, accuracy of identity recognition is improved.