Terminal Image Recognition Model for Local Labeling

Solution Overview

Problem

The increasing number of terminals in image recognition systems leads to a high workload for servers, reducing processing efficiency and compromising user privacy, as images must be uploaded for label information labeling.

Innovation Solution

Implementing an image recognition method on terminals using a built-in image recognition model to obtain and store object category information locally, reducing the need for server-based processing and enhancing privacy protection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If images are uploaded to a server for label information labeling, then label information can be obtained, but server workload increases and processing efficiency decreases

Engineering Contradiction:

Improvelabel information accuracyVSAvoidprocessing efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The terminal performs image recognition and label information generation autonomously using a locally deployed recognition model, eliminating the need to upload images to the server. The terminal independently completes the entire process of image analysis and label generation, thereby reducing server workload and improving processing efficiency while maintaining label accuracy

Inventive Principle:

Principle #25Self-service

2Reliability

If images are uploaded to a server for processing, then label information can be obtained, but user privacy is compromised

Engineering Contradiction:

Improvelabel information accuracyVSAvoidprivacy protection

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The terminal performs image recognition and label information generation autonomously using a locally deployed recognition model, eliminating the need to upload images to the server. The terminal independently completes the entire process of image analysis and label generation, thereby reducing server workload and improving processing efficiency while maintaining label accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The recognition model is extracted from the server environment and deployed directly on the terminal device. This extraction allows the terminal to perform image recognition locally without transmitting images to the server, thereby protecting user privacy while maintaining recognition accuracy

Inventive Principle:

Principle #2Taking out (Extraction)

3Adaptability or versatility

If a server processes images for label information, then centralized control is maintained, but system complexity increases

Engineering Contradiction:

Improvesystem flexibilityVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system is segmented into independent terminal units, each capable of autonomous image recognition. This segmentation distributes the recognition functionality from a centralized server to individual terminals, reducing system complexity while maintaining flexibility through modular, independent operation of each terminal

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11804053B2Image recognition method and terminal

Publication Date: 2023.10.31 HUAWEI TECH CO LTD

AI summary

An image recognition method and a terminal, where the method includes obtaining, by the terminal, an image file comprising a target object, recognizing, by the terminal, the target object based on an image recognition model in the terminal to obtain object category information of the target object, and storing, by the terminal, the object category information as first label information of the target object. Hence, image recognition efficiency of the terminal can be improved, and privacy of a terminal user can be effectively protected.