Bibliographic citations
Carpio, J., Calderon, J. (2022). Sistema electrónico prototipo basado en algoritmos de procesamiento digital de imágenes orientado al control de calidad textil en el proceso de tejido [Tesis, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/667480
Carpio, J., Calderon, J. Sistema electrónico prototipo basado en algoritmos de procesamiento digital de imágenes orientado al control de calidad textil en el proceso de tejido [Tesis]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2022. http://hdl.handle.net/10757/667480
@misc{renati/404479,
title = "Sistema electrónico prototipo basado en algoritmos de procesamiento digital de imágenes orientado al control de calidad textil en el proceso de tejido",
author = "Calderon Gutierrez, Jorge Ismael",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2022"
}
The present work proposes the creation of a prototype electronic system based on image processing algorithms for the quality control of flat fabric that will allow the detection of three types of faults (holes, packages and stains) in textile processes. This prototype system is created in such a way that it is adaptable and scalable to textile companies. The implementation of this project mitigates the risk posed by the human factor in the quality control process, which it this the problem of the client by the less reliability which generate the mental and physical fatigue the operator than an automatic electronic system that does not have a precision- fatigue in its operation. Currently there are solutions at very high cost for small and medium-sized textile companies. The scientific literature proposes some solutions with processing algorithms, deep-learning and convolutional networks. This solution proposes a fusion between hardware and software. The hardware includes a small board computer and the elements for the control of the camera, lighting and the engine while the software is based on Python, which is an open source language and is composed of digital image processing algorithms of low computational load of the images obtained, which will be manipulated through segmentation and thresholding, which will emit results quickly with a high reliability for the location and recognition of the faults in the fabrics of different shades and colors. The validation and feasibility of the project was carried out by 1 specialist engineer in the textile sector, obtaining a reliability percentage greater than 92% in the detection of textile defects with an average detection speed of 15.68 m/min.
This item is licensed under a Creative Commons License