Non-destructive Classification of woven fabrics using Machine Learning

The project is about classification of woven fabrics such as plain, satin and twill satin fabrics based on a machine learning model for which we will be using image processing techniques. The novelty of this project is that it will provide non-destructive way for the fabric classification apart from

Project Title

Non-destructive Classification of woven fabrics using Machine Learning

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

The project is about classification of woven fabrics such as plain, satin and twill satin fabrics based on a machine learning model for which we will be using image processing techniques. The novelty of this project is that it will provide non-destructive way for the fabric classification apart from the conventional techniques.

With the advancement in computer technology, new automatic and efficient ways are required for problem solving and making work of human easier. In today’s textile industry, fabric classification is done manually which require extensive human labor and considerable efforts. This novel method of woven fabric classification will be really helpful in customs department and other such places where typical methods are used for fabric classification.

Project Objectives

Hardware:
• Raspberry Pi hardware along with a mini digital microscope will be used that will act as a hand-held device.
Software:
• Classify woven fabrics using Image Processing Techniques.
• Python IDE will be used to develop the algorithm. We willthen train and test Machine Learning model using SVM for fabric classification.

Project Implementation Method

We have used image processing techniques to obtain the required features of fabric image. 
Images are obtained using a scope of upto 800X magnification. 
Dataset is prepared using these fabric images and data augmentataion techniques are applied to further increase the dataset size. These fabric images are further processed using SVM. 
Furthermore, a machine learning model will be trained and tested such that it will provide results in real-time

Benefits of the Project

• Real time hand held device for fabric classification.
• Automatic classification of woven fabrics in customs department.
• The developed handheld hardware will be useful for cloth merchants as well.

Technical Details of Final Deliverable

The device will comprise of mini portable digital microscope with an optical zoom of up to 800X and an 8-megapixel (MP) camera fixed on translucent off-white acrylic sheets. 

The background of the acrylic is illuminated using LEDs, intensity of which can be controlled manually as per need. 
This is done so as to clearly observe and distinguish between the warp and weft patterns at a set intensity of background light for every fabric sample. 
The light passing from the spacing between the warp and weft threads is now visible in the camera. The algorithm will thus classify the fabric type as plain, twill or satin on these basis.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Others

Other Industries

Core Technology

Others

Other Technologies

Sustainable Development Goals

Decent Work and Economic Growth, Reduced Inequality

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com