Improving Detection and Counting Accuracy of Humans Using Convolutional Neural Networks Compared to Principal Component Analysis

To enhance human detection and counting accuracy by employing a novel convolutional neural network (CNN) instead of the Principal Component Analysis (PCA) algorithm. The two categories in this study are Principal Component Analysis (PCA) and Convolutional Neural Networks (CNN). The study parameters include alpha value 0.05, beta value 0.2, and power value 0.8, with 136 samples in each group. Additionally, utilizing various sample sizes, their accuracy comparisons are made with one another. In Human Detection and Counting, the Novel Convolutional Neural Network outperforms the Principal Component Analysis Algorithm by 91.39 %, with a precision of 75.65 %. When it comes to identifying and counting humans, the CNN model performs noticeably better than the PCA.