An Efficient Data Analytics and Optimized Algorithm for Enhancing the Performance of Image Segmentation Using Deep Learning Model

Image segmentation is the key topic in computer vision and image processing with applications like robotic perception, scene understanding, video surveillance, compression, medical analysis, augmented reality among many others. There are numerous algorithms developed literature for segmentation. This paper provides a broad spectrum of pioneering works instance semantic level mask Region based Convolution Neural Network Monarch butterfly Optimization (RCNN-MBO) architecture. The system initially constructed Python environment images people animals being input. Remove unnecessary data from gathered datasets during pre-processing stage. Next, use stochastic threshold function to segment image. Then update segmented into designed model detecting classifying group images. main goal approach attain accurate prediction results also improve performance by attaining better results. To enhance performance, two activation functions were used MBO fitness updated classification layer. It improves takes less time detect classify Finally, experimental outcomes show reliability other conventional techniques terms accuracy, precision, sensitivity, specificity, F-measure, error rate, computation time.