Automated Image Analysis System for Homogeneity Evaluation of Nuclear Fuel Plates

The main aim of this work is to design an automated image analysis system developed for inspection of fuel plates manufactured for the operation of ETRR-2 of Egypt. The proposed system aims to evaluate homogeneity of the core of the fuel plate. A vision system has been introduced to capture images for plates to be characterized and software has been developed to analyze the captured images based on the gray level co-occurrence matrix (GLCM). The images are digitized using digital camera, and it is common practice to adopt a preprocessing step for the images with a special purpose of reduction/eliminating the noise. Textural features of angular second moment, homogeneity, entropy, contrast and inverse difference moment were extracted from images of fuel plates and used in back propagation neural network (BPNN). The back propagation neural network is used to classify the fuel plates as a "regular structure" or "defect". The experimental results show that higher classification rate and The BPNN model was able to discriminate between the regular structure and the defect classes with good classification accuracy