A Novel Defocused Image Segmentation Method Based on PCNN and LBP

The defocus blur concept adds an artistic effect and enables enhancement in the visualization of image scenery. Moreover, some specialized computer vision fields, such as object recognition or scene restoration enhancement, might need to perform segmentation separate blurred non-blurred regions partially images. This study proposes a sharpness measure comprised Local Binary Pattern (LBP) descriptor Pulse Coupled Neural Network (PCNN) component used implement robust approach for segmenting in-focus from out focus sections scene. proposed is very sense that parameters model can be modified accommodate different settings. presented metric exploits fact that, general, local patches blurry have less prominent LBP descriptors than non-blurry regions. combines this with PCNN algorithm; images are segmented along clear edges objects. has been tested on dataset 1000 defocused eight state-of-the-art methods. Based set evaluation metrics, i.e., precision, recall, F1-Measure, results show algorithm outperforms previous works terms accuracy efficiency improvement. also uses other parameters, Accuracy, Matthews Correlation Coefficient (MCC), Dice Similarity (DSC), Specificity, assess better obtained by our proposal. we adopted fuzzy logic ranking scheme inspired Evaluation Distance Average Solution (EDAS) technique interpret integrity. experimental outputs illustrate referenced methods optimizing quality reducing computational complexity.