Realtime Visual Object Recognition using Support Vector Machine comparing with K- Nearest Neighbor algorithm for improving accuracy
DOI:
https://doi.org/10.47750/pnr.2022.13.S04.097Keywords:
Novel Support Vector Machine, Machine Learning, Object Recognition, K- Nearest Neighbor, Bounding boxes, Feature extraction.Abstract
Aim: The aim of this research is to recognise objects by using machine learning algorithms from the images with improved accuracy. Materials and Methods: A total of 104 samples are there for the two groups. Novel Support Vector Machine is considered as group 1 and K- Nearest Neighbor Algorithm is considered as group 2. Group 1 consists of 52 samples and Group 2 also consists of 52 samples and the G power is 80%.
Results: The accuracy for the Novel Support Vector Machine algorithm (92%) is more than that of the K- Nearest Neighbor algorithm (83%). The mean accuracy detection is ±2SD and the significance value is 0.000 (p<0.01) which shows the hypothesis is correct and it is carried out using an independent sample T
test.
Conclusion: Hence, the accuracy of Novel Support Vector Machine is found to be 92% which is more than the accuracy of the K- Nearest Neighbor algorithm which is 83%.
