Scene Semantic Recognition Based on Modified Fuzzy C-Mean and Maximum Entropy Using Object-to-Object Relations

With advances in machine vision systems (e.g., artificial eye, unmanned aerial vehicles, surveillance monitoring) scene semantic recognition (SSR) technology has attracted much attention due to its related applications such as autonomous driving, tourist navigation, intelligent traffic and remote sensing. Although tremendous progress been made visual interpretation, several challenges remain (i.e., dynamic backgrounds, occlusion, lack of labeled data, changes illumination, direction, size). Therefore, we have proposed a novel SSR framework that intelligently segments the locations objects, generates Bag Features, recognizes scenes via Maximum Entropy. First, denoising smoothing are applied on data. Second, modified Fuzzy C-Means integrates with super-pixels Random Forest for segmentation objects. Third, these segmented objects used extract Features concatenate different blobs, multiple orientations, Fourier transform geometrical points over An Artificial Neural Network using patterns Finally, labels estimated Entropy model. During experimental evaluation, our system illustrated remarkable mean accuracy rate 90.07% MSRC dataset 89.26% Caltech 101 object recognition, 93.53% Pascal-VOC12 respectively. The should be applicable various emerging technologies, augmented reality, represent real-world environment military training engineering design, well entertainment, eyes visually impaired people monitoring avoid congestion or road accidents.