Publication Date: 2023/06/17
Abstract: The purpose of image segmentation methods is to create analytically relevant subsets of an input image. Segmentation is typically driven by the input information and a precondition on the search area; the latter is useful when the images are damaged or contain artifacts due to limitations in the image collection technique. It is possible for image segmentation techniques to make use of prior knowledge in order to deliver outcomes that are more accurate and credible. The method known as event-based imaging makes it possible to recognize occurrences in a way that is both efficient and helpful by using the medium of pictures. This is a very sophisticated system that requires the cognitive categorization of the components in the picture as well as the proper recognition of the event. For the purpose of event-based imaging, there have already a great number of studies and investigations conducted, all of which have been created with this particular objective in mind. However, it has come to everyone's attention that the bulk of the prevalent researches are unable to independently conduct the event identification with a significant degree of accuracy. Therefore, to provide a solution to this problem this research devises an effective methodology that utilizes Image normalization, image segmentation and Channel Boosted Convolutional Neural Networks to achieve event recognition.
Keywords: Event based imaging, Image normalization, Image segmentation, Channel Boosted Convolutional Neural networks, Decision Tree.
DOI: https://doi.org/10.5281/zenodo.8049693
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23MAY720.pdf
REFERENCES