Automatic Detection of Edges in Handmade Embroidery Patterns

Abstract

Aims/ Objectives: The study examined the specific characteristics responsible for the recognition of edges in handmade embroidery patterns, designed a computational model for the process, implemented the model and evaluated its performance. This is with the view to detecting the edges of handmade embroidery patterns in the context of computational modeling. Study Design: Computational modeling. Place and Duration of Study: Department of Computer Science and Engineering and Department of Fine and Applied Art, between February 2016 and May 2017. Methodology: Samples of hand embroidery patterns were collected through embroiderer shop in Ìbàdàn, Òsogbo, and Ilé-Ifè and the collected samples were pre-processed and the edges of the patterns were detected using cellular automata (CA) and cellular learning automata (CLA). The performance of the system was evaluated in terms of computing time, Mean Square Error (MSE) and Peak Signal-to-Noise Ratio (PSNR). Results: The result obtained from all the experiments carried out showed that the Cellular Edge Detection (CED) algorithm has lower value in terms of MSE, a higher value in terms of PSNR and lesser computational time as compared to the standard edge detection algorithm. The automatic detection of edges showed that the complex stitches of handmade embroidery patterns are amenable to computational rendering through efficient and effective techniques. Conclusion: This study will enhance the performance of the edge detection techniques employed in pattern recognition and computer vision applications.