Ontology-based video retrieval using modified classification technique by learning in smart surveillance applications

Abstract

With the rapid advancements in communication and multimedia computing technology, multimedia information, particularly video data, has recently become disproportionately accessible. Video is widely used in a variety of applications, making efficient management and retrieval of the expanding volume of video data critical. To manage such tasks, include automated recognition of the image queries in video retrieval with a reduced degree of system memory usage, spot detection, maintenance time, identification of exact duplicate videos, and so on. This research determination demonstrates the Modified R-Ratio with Viola-Jones Classification Method (MRVJCM) with the relevance of developing techniques and algorithms for automatic recognition of image queries. The R-Ratio Viola–Jones framework has two distinguishable feature maintenance processes, such as Feature Selection and Integration and Feature Cascading. These two distinct features are applied to the motion features (texture, emotions, elements, and shape) within the three features. The proposed techniques and their relevant algorithms are used to retrieve the most accurate videos and to assure the mathematical operation of video retrieval in the operation of the comparable protected system. As a result, the proposed MRVJCM achieves 98% of accuracy, 93% of precision, 92% of recall, and 42% of RMSE.