Baggage Image Classification Using AI Color Mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing baggage handling systems face challenges in accurately identifying and categorizing bags due to subjective human descriptions, inconsistent labeling, and variations in color perception, leading to high mishandling rates and passenger complaints.

Innovation Solution

An AI-based system using multiple computer vision models processes a single image of a bag to determine type, material, and external elements, combining outputs for improved accuracy, and employs a color mapping process using HSV definitions and machine learning to objectively categorize colors, reducing the need for additional infrastructure and human intervention.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional color determination algorithms using distance functions in 3D color space are used, then color categorization can be performed, but the results do not match human color perception due to individual variations in perception

Engineering Contradiction:
Improvecolor categorization accuracyVSAvoidhuman perception alignment
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent transforms the continuous 3D color space parameters (RGB, HSV) into discrete color category labels through machine learning classification. This parameter transformation enables the system to output colors that align with human categorical perception rather than mathematical distance metrics.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical distance-based color determination algorithm with a machine learning-based classification system. This substitution allows the system to learn human color perception patterns from training data rather than relying on mathematical distance functions that do not reflect human cognition.

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

2Measurement precision

If multiple images are captured to improve classification accuracy, then more information is available, but additional infrastructure and complexity are required

Engineering Contradiction:
Improvebag classification accuracyVSAvoidinfrastructure requirements
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent enhances the information content of a single image by extracting multiple features (color histograms, texture features, shape descriptors) from different parameter spaces. This allows the system to achieve multi-image-level classification accuracy using only a single captured image.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent adds dimensional depth to single-image analysis by incorporating multiple feature dimensions (color, texture, shape, spatial relationships) and processing stages. This dimensional enrichment compensates for the lack of multiple images while maintaining high classification accuracy.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Ease of operation

If subjective human descriptions of bags are used for identification, then passenger input is captured, but matching descriptions to specific bags becomes difficult and error-prone

Engineering Contradiction:
Improvepassenger input capabilityVSAvoidbag identification accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent creates an objective visual copy (image) of the bag and processes it through computer vision algorithms to generate structured classification data. This visual copy serves as an objective reference that can be accurately matched against passenger descriptions, eliminating the ambiguity of purely subjective verbal descriptions.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent introduces computer vision classification as an intermediary between the physical bag and the passenger's description. This intermediary transforms the physical bag into structured visual features that can be objectively compared with description keywords, improving matching accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

4Adaptability or versatility

If manual baggage identification processes are used, then flexibility is maintained, but time consumption and labor costs increase significantly

Engineering Contradiction:
Improveprocessing flexibilityVSAvoidbaggage handling speed
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent enables the baggage identification system to perform classification and matching operations autonomously without human intervention. The computer vision system automatically extracts features, categorizes bags, and matches them against descriptions, eliminating manual labor while maintaining adaptability through programmable algorithms.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual human inspection and identification processes with automated computer vision and machine learning systems. This substitution dramatically increases processing speed while maintaining flexibility through software-based classification rules that can be adapted to different scenarios.

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

Data Source

PatentEP3716148B1Item classification system, device and method therefor
Publication Date: 2026.04.22 SITA INFORMATION NETWORKING COMPUTING UK LTD
  • EP3716148B1 patent drawingFigure 1
  • EP3716148B1 patent drawingFigure 2
  • EP3716148B1 patent drawingFigure 3

AI summary

An image processing system for categorising the colour of an item is disclosed. The system comprises processing means configured to: process an image of an item to extract a portion of the image where the item is located; determine a first average colour value of a plurality of colour values associated with the portion of the image where the item is located; map the average colour value to one of a plurality of predetermined colour definitions based on a plurality of colour ranges associated with each colour definition; and categorise the colour of the item according to the mapping.