Automatic Recognition of Metal Fiber Per Unit Area for Electromagnetic Shielding Fabric Based on Computer Image Analysis

Metal fiber content is often a measured parameter for electromagnetic shielding fabric (ESF). A commonly used method is combustion measuring, but measuring speed was slow and measured fabric damaged. This study proposes a new method based on computer image analysis for recognition of metal fiber content per unit area (MFCPUA) of the ESF, which aims at analyzing the MFCPUA without damage and providing a basis for the shielding performance evaluation of the ESF. Local region images of garment or fabric are obtained using high definition shooting system to build a gray matrix model which can describe the image. A recognition algorithm for fabric density based on gray extreme judgment is then given to construct a computation for the MFCPUA. The recognition results obtained with the proposed method is compared with the experimental results from manual combustion measuring, and the error reason and the application are also analyzed. Results of experiments and analyses show that the proposed method can identify the local fabric density with lossless and accurately calculate the MFCPUA, which provide a new method for electromagnetic shielding performance evaluation of the fabric and garment by computer technique.