STTR Phase I: IntelliStitch AI: Intelligent Computerized Embroidery Design Automation for the Textile Industry

Soft Sight, Inc., Vestal NY

Investigators

Abstract

This Small Business Technology Transfer Research (STTR) Phase I project builds upon the technology and success of the company's software product, one of the industry's first embroidery design automation systems. It provides the textile industry with simplified mechanisms for converting scanned artwork into high quality embroidery design data. This data is then utilized by commercial sewing equipment to produce the embroidered artwork that has become quite common on all types of garments and woven goods. Unfortunately, embroidered artwork is often quite expensive to produce. In many cases, it may substantially exceed the costs of the actual garments being imprinted. These costs arise from a variety of factors including an embroidered design's size, layout, and complexity. Well-designed embroidered artwork permits efficient production with high yields (i.e. causing very few defective items to be produced). Automating this design creation process provides additional benefits by eliminating the time consuming manual process that must otherwise be undertaken by a human expert. With these factors as the primary motivation, this research will investigate advanced artificial intelligence and machine vision mechanisms, such as neural nets and structural indexes, to substantially improve capabilities and performance. The advantages of a robust automation system to the textile industry are quite substantial. Creating sophisticated embroidery designs is a tedious, time-consuming activity requiring the skills of a human expert (called a digitizer). Even after this process has been completed, any miscalculations by the digitizer could substantially impede production on machinery. A well-designed expert system could inevitably eliminate these costs and perhaps even provide a level of quality that is not achievable by its human counterparts. Additionally, this research may also have broad application within other fields such as document processing or other areas where image understanding and interpretation is important.