IC testing using thermal image based on intelligent classification methods

The goal of this thesis is to propose an algorithm which would can locate the defect IC on the PCB during their manufacturing phase based on a thermal image. A 3-dimensional PCB finite-element model is developed to estimate the temperature profile of stacked ICs. Image processing by noise removing and region of interest segmentation are applied. Two sets of feature extraction are presented; first-order histogram features and Gray Level Co-occurrence Matrix (GLCM) features. The Principle Component Analysis (PCA) method is applied to decrease the feature's extractions into smallest uncorrelated input. Three main intelligent techniques; Multilayer Perceptron (MLP), Support Vector Machine (SVM), and Adaptive Neuro-Fuzzy Inference System (ANFIS) are used to classify the thermal conditions of ICs into normal and faulty status. On validation, the proposed approach applies to do thermal testing on Arduino UNO. The experimental evaluation is performed to detect the fault condition on the real time operating PCB.