Knit Fabric Database Search Using Visual Construction Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods of organizing fabric data do not allow for easy searching, organizing, or matching of fabrics with similar construction characteristics, leading to inefficiencies in fabric design and production, as they lack specificity and require excessive collaboration and trial and error.
Innovation Solution
A digital image-based system utilizing AI and machine learning to analyze and categorize knit fabrics by construction attributes, enabling a searchable database and interface for efficient fabric selection and reproduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physical fabric samples are stored in sample books or libraries and physically circulated, then designers can review actual material quality and texture, but collaboration becomes difficult for designers at separate or remote locations and time is lost in physically circulating samples
Solution Approach 1:
The patent creates high-resolution digital images and 3D virtual models as copies of physical fabric samples. These digital representations preserve all visual and textural characteristics while eliminating the need for physical sample circulation. Designers can access and review fabric details remotely through the digital library, maintaining full inspection capability without geographical or temporal constraints.
Solution Approach 2:
The patent replaces the mechanical system of physical sample handling, storage, and circulation with an automated digital imaging and database system. Cameras and scanners capture fabric properties, which are then stored and retrieved through computer interfaces, eliminating manual sample transport and physical library management while improving accessibility and collaboration efficiency.
2Ease of operation
If basic fabric information such as color, generic construction tags, and fabric type is provided, then simple identification is possible, but the information lacks the specificity needed for future production or reproduction of the fabric
Solution Approach 1:
The patent segments fabric information into multiple hierarchical levels: basic identification data (color, fabric type), detailed construction attributes (stitch type, yarn composition, weave pattern), and comprehensive production specifications (machine settings, tension parameters, finishing processes). This segmentation allows users to access only the detail level needed for their specific task, maintaining simplicity for basic identification while preserving complete specificity for production reproduction.
Solution Approach 2:
The patent transforms fabric characteristics into quantifiable parameters and structured data fields that can be precisely measured, stored, and retrieved. Construction attributes are converted from qualitative descriptions to standardized parameters that capture the complete technical specification needed for fabric reproduction, enabling both simple browsing and detailed manufacturing guidance.
3Ease of operation
If textile manufacturers provide basic fabric information, then initial fabric identification is possible, but excessive collaboration and trial and error are required to determine the construction of a fabric
Solution Approach 1:
The patent performs preliminary analysis and documentation of complete fabric construction data during the initial sampling phase. All construction attributes, material specifications, and production parameters are captured and stored in the database before any collaboration or production attempts begin. This preliminary action eliminates the need for subsequent trial and error by providing complete construction information upfront to all collaborators.
Solution Approach 2:
The patent implements a feedback system where the digital database provides immediate access to complete fabric construction information, allowing designers and manufacturers to verify specifications without iterative sampling. The system feedback loop eliminates productive delays by providing instant access to authoritative fabric data, reducing collaboration cycles from multiple iterations to single-step verification.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
Systems and methods are discussed herein and include a method of providing for search and review of an electronic database of knit fabric samples, the method including: providing a plurality of digital images of a plurality of knit fabrics; organizing the plurality of digital images in the database based upon corresponding visual cues extracted using an image analysis performed on each of the plurality of digital images, wherein the visual cues are construction attributes of the plurality of knit fabrics; displaying one or more of the plurality of digital images for review; and providing for search of the database using one or more of the visual cues.