Fault detection using thermal image based on soft computing methods: Comparative study

A 3D PCB model is designed to detect the faulty IC based on IR images. Three classification techniques are used in this work; MLP, SVM, and ANFIS. The dimensions, and physical properties are selected to generate the thermal profile for a numerical population of data sets, and then histogram features are computed which are used as inputs into a classifier model after minimizing the number of effective features by the PCA.92