Applying Template Matching Technique for Distortion Removing from Photography

• Published in: 2022 8th International Conference on Contemporary Information Technology and Mathematics (ICCITM) pp. 347 - 351
• Main Authors: Alobaydi, Esraa Khalid Ahmed, Abdullah, Omar Muayad
• Format: Conference Proceeding
• Language: English
• Published: IEEE 31-08-2022
• Subjects: Distortion; distortion removing; Image processing; Indexes; Mathematical models; Measurement; Photography; Position measurement; Size measurement; Template matching
• DOI: 10.1109/ICCITM56309.2022.10031686

The paper aims to remove distortion from analog media like old photographs due to several reasons like shredding and using multiple template matching techniques. First of all, is to convert the determined photograph into its digital representation to get the pixels' intensities, then we extract the distorted templates to manipulate them by scanning the templates one by one on the whole image to determine the template position in the image, this achieved by applying Structural Similarity Index Measurement SSIM metric between the determined distorted template and the corresponding template in the whole image, this process is called matching step, then after determining the distorted template, now we will apply the proposed algorithm (equations) on that template to remove the distortion, also we applied a comparison using both SSIM and Mean Squared Error MSE metrics to determine the optimal template size depending on the following factors processing efficiency and execution speed.