An Integrated Statistical Process Monitoring and Fuzzy Transformation Approach to Improve Process Performance via Image Data

Due to the increased volume of data generated in the form of images, the role of various image processing and monitoring methods has become vital in improving quality. Monitoring image data requires development and implementation of effective data reduction methods. The extracted data from various image types are considered a rich and valuable source of information in statistical process monitoring (SPM). In this paper, we propose a new method to monitor extracted data from 2-dimensional (2D) grayscale images based using an integrated approach based on fuzzy transform (F-transform) and generalized likelihood ratio (GLR) test. F-transform is applied first to reduce the image dimension effectively and then the extracted data are used for statistical process monitoring utilizing a generalized likelihood ratio (GLR) control chart. The proposed approach can help practitioners to identify fault location and change point in the process. A case study based on real image data in tile industry along with numerical examples are also presented to evaluate the performance of the proposed method. Results indicate that the proposed approach has satisfactory performance from statistical point of view for detecting variations in processes.