Efficient System for Criminal Face Detection Technique on Innovative Facial Features To Improve Accuracy Using LBPH In Comparison With CNN
DOI:
https://doi.org/10.47750/pnr.2022.13.S04.085Keywords:
Local Binary Pattern Histograms, Convolutional Neural Network, Image Classification, Facial Image, Innovative Facial Features, Histograms, Face Recognition, Machine Learning.Abstract
Aim: The objective of this study is to use face recognition technology to identify the suspects through facial biometrics which helps in identification of criminals faster based on innovative facial features. Innovative Local Binary Pattern Histograms algorithm is used for the face recognition method.
Materials and Methods: Face detection for identifying the criminals is performed using Innovative Local Binary Pattern Histograms (N=10) was iterated 10 times for efficient and accurate analysis based on labeled data with G power in 80% and threshold 0.05%, CI 95% mean and standard deviation. The split size of training and testing of 70% and 30% respectively.
Results: It is found that the accuracy of LBPH is 90.80% which is higher than the CNN model 90.30% and attained the significance value of p = 0.0303 (p<0.05).
Conclusion: For the face identification purpose Local Binary Pattern Histograms algorithm is preferred than Convolutional Neural Network algorithm.
