Computational Intelligence-Based Harmony Search Algorithm for Real-Time Object Detection and Tracking in Video Surveillance Systems

Recently, video surveillance systems have gained significant interest in several application areas. The examination of sequences for the detection and tracking objects remains a major issue field image processing computer vision. object process includes extraction moving from frames continual over time. latest advances computation intelligence (CI) techniques become popular In this aspect, study introduces novel computational intelligence-based harmony search algorithm real-time (CIHSA-RTODT) technique on systems. CIHSA-RTODT mainly focuses detecting that exist frame. incorporates an improved RefineDet-based module, which can effectually recognize multiple addition, hyperparameter values RefineDet model are adjusted by use Adagrad optimizer. Moreover, (HSA) with twin support vector machine (TWSVM) is employed classification. design optimal feature HSA to appropriately adjust parameters involved TWSVM shows novelty work. A wide range experimental analyses carried out open access dataset, results inspected ways. simulation outcome reported superiority other existing techniques.