Introduction: There is a belief in Islam that performing ablution before each prayer usually enhances the brightness of faces. In line with this assertion, there has not been proper scientific proof in determining facial conditions of Muslims in respect of the ablution effect to the brightness of their faces.
Aims: To examine, investigate, and determine the strengths and weaknesses of the available machine learning algorithms for facial image analysis with respect to the effects of ablution on the human visage.
Materials and Methods: The Google Scholar repository yielded 265 publications, all of which were thoroughly reviewed. After our predetermined inclusion and exclusion criteria were applied, only 48 papers were chosen for analysis in this study.
Results: Results show that LBP, PCA, CNN, and SVM are the most frequently used algorithms for facial image categorization systems, and the accuracy of SVM ranged from 94.5% to 96.83%, and CNN returned an accuracy of 98.3%. Results also show there is a correlation between the size of input data and the accuracy of machine learning algorithms.
Conclusion: This suggests that the Support Vector Machine (SVM) algorithm returned a higher accuracy of 96.83% in classification, especially when used with fewer datasets, and CNN worked robustly on larger datasets with an accuracy of 98.3%.