Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
Aim: To detect the damaged number plates with ROI (Region of Interest) using advanced Convolutional Neural Networks (CNN) and comparing them with NBC (naive bayes network). Materials and Methods: Classification is performed by Novel Convolutional Neural Networks (CNN) (N=10) over (NBC) naive bayes network (N=10). Sample size is calculated using Gpower with a pretest power as 0.8 include alpha value 0.05, beta value 0.2. Results: Mean accuracy of the Novel Convolutional Neural Networks is (96.40%) is high compared to naive bayes network of (93.30%). Significance value for accuracy and loss is 0.01 (p>0.05) Conclusion: The Mean Accuracy of damaged number plate in Novel Convolutional Neural Networks is better than the naive bayes network.