Technical Application of Dynamic Visual Recognition System in Textile Industry Advertising and Brand Communication

Abstract

With the rapid development of intelligent multimedia communication and immersive information delivery, dynamic visual perception technologies are becoming increasingly important for digital advertising systems and next-generation interactive communication applications. To overcome the limited engagement and weak brand memorability of conventional static textile advertisements, this study proposes a dynamic visual recognition system integrating YOLOv8-based object detection, convolutional neural network (CNN) feature extraction, and augmented reality (AR) rendering. The framework first detects fabric textures, garment outlines, and brand logos in real time, and subsequently constructs multidimensional representations of brand elements through hierarchical feature modeling and fusion. The extracted semantic features are then incorporated into an AR-driven interactive presentation mechanism to generate immersive advertising experiences. Experimental results demonstrate that the proposed system significantly improves user attention, brand recall, and interactive engagement, with brand logo gaze duration reaching 2450 ms, free recall increasing to 78.5%, and participation in fabric-detail exploration achieving 92.5%. By combining intelligent visual recognition with dynamic information presentation, the proposed approach provides an effective solution for enhancing advertising communication and offers valuable technical insights for vision-assisted multimedia transmission and intelligent electromagnetic information interaction systems.