Volume 20 No 13 (2022)
 Download PDF
Strömberg Wavelet Transform Feature Extracted based Deep Belief Learning Classification for Pedestrian Detection
H. Sivalingan, Dr. N. Anandakrishnan
Video surveillance system is a network of cameras, monitor/display unit and recorder. Surveillance task plays an essential part for finding the abnormal scenarios. Pedestrian detection is a key component for intelligent transport system and driver assistance system. But, the existing algorithms were inadequate for small-scale pedestrian detection. In order to address these problems, Strömberg Wavelet Transform Feature Extracted based Deep Belief Learning Classification (SWTFEDBLC) Technique is introduced. Initially, the number of images is collected from the input database at the input layer. Consequently, preprocessing of input image is carried out in SWTFE-DBLC Technique to remove the noisy pixels from input image in hidden layer 1. After that, the features from the input images are extracted through transformation method in hidden layer 2. Then, the feature mapping is performed through determining the part score in the hidden layer 3. Lastly, the activation function is used in the output layer for performing efficient crime detection. By this way, SWTFE-DBLC Technique performs efficient pedestrian detection with higher accuracy and lesser time consumption. Results illustrates that the proposed SWTFE-DBLC Technique significantly improves true positive rate with minimal false positive rate and crime detection time with respect to image size and number of images
Video surveillance system, pedestrian detection, preprocessing, transformation, deep belief learning classification
Copyright © Neuroquantology

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Articles published in the Neuroquantology are available under Creative Commons Attribution Non-Commercial No Derivatives Licence (CC BY-NC-ND 4.0). Authors retain copyright in their work and grant IJECSE right of first publication under CC BY-NC-ND 4.0. Users have the right to read, download, copy, distribute, print, search, or link to the full texts of articles in this journal, and to use them for any other lawful purpose.