Intelligent dandelion optimizer-driven weighted support vector machine: Design of a spatial visual feedback system for augmented reality art installations

• Developed SVFS with IDO-WSVM and projector AR for real-time adaptive visual feedback. • Achieved 97.56% label visibility prediction, 8.22% higher than baseline WSVM. • Enabled low-latency feedback with 94 ms average adjustment for immersive experience. • Dimensionality reduction cut processing time from 120 ms to 67 ms efficiently. • User evaluations showed high satisfaction: spatial responsiveness 2.25, speed 2.15. Augmented Reality (AR) art installations provide immersive, interactive experiences, yet ensuring clear and responsive visual communication in dynamic spatial environments remains challenging. This study proposes a Spatial Visual Feedback System (SVFS) that dynamically adjusts labels, linkage lines, and visual cues based on real-time spatial parameters such as user viewpoint, surface type, and occlusion. The system integrates projector-based Spatial AR with an Intelligent Dandelion Optimizer-driven Weighted Support Vector Machine (IDO-WSVM) to predict label visibility. A dataset was collected from 45 visually healthy participants, each performing 160 projection tasks across four tabletop configurations, capturing spatial features including view angle, distance, occlusion percentage, and surface type, along with vocal feedback on label clarity. The IDO-WSVM model achieved 97.56% visibility prediction accuracy, outperforming baseline WSVM by 8.22%, and enabled real-time visual feedback with an average latency of 94 ms. Feature dimensionality reduction further improved computational efficiency, reducing processing time from 120 ms to 67 ms without compromising accuracy. The novelty of this research lies in combining evolutionary optimization with weighted SVM to provide context-aware, adaptive visual feedback in AR environments, a capability absent in prior art-focused AR systems. User evaluations demonstrated high satisfaction in spatial responsiveness (2.25) and feedback speed (2.15), confirming enhanced engagement and immersive experience. The proposed SVFS establishes a scalable, intelligent framework for AR art installations, capable of accommodating diverse surfaces, viewpoints, and user interactions, advancing both the technological and experiential aspects of spatial AR in artistic contexts.