多维感官的动态信息数字媒体传递：从数据流变到感知重构

From Data Streaming to Perceptual Reconfiguration: Transmitting Multidimensional Sensory DynamicInformation via Digital Media

关键词:  风景园林  多模态感知  实时渲染  数字孪生  PLS-SEM/线性混合模型  感受计量

Key words:  landscape architecture  multimodal perception  real-time rendering  digital twin  PLS-SEM/linear mixed model  perceptual measurement

摘要:

多媒体技术与数字孪生环境迅猛发展，但针对多模态感知的量化评估与反馈机制尚不完善。为此，提出了一个4层感受计量模型，基于DLES指标构建了 用户体验评分体系，并设计了实时渲染与数据反馈的闭环机制。研究成果包括建立理论模型、开发感知评价工具，并通过PLS-SEM和线性混合模型等统计方法对 实证案例进行了验证。旨在从传播学视角，为理解数字媒介中多维感官信息的流变(flux)与感知重构(reconfiguration)机制提供理论框架与分析方法。

Abstract:

Emerging multimedia and digital-twin technologies are increasingly applied in art, education, tourism, and landscape design, creating highly interactive immersive environments that engage multiple human senses. Digital media now support simultaneous visual, auditory, and haptic stimuli while allowing users to respond in real time, transforming communication from linear broadcast to a dynamic, bidirectional sensory dialogue. Conceptually, this study treats digital content as a continuous data stream and highlights perceptual reconfiguration - the active, experience-driven construction of meaning from converging multisensory inputs. Despite these advances, systematic methods to quantify multimodal user experience and to feed it back into design optimization remain lacking. This gap is especially pertinent for interactive museum exhibits, VR education modules, and virtual therapeutic environments, where orchestrating audio-visual-haptic cues is key. To address this need, we propose a unified framework that measures users' sensory responses and loops these insights back into adaptive landscape media design. We articulate a novel four-layer sensoriometric model. The lower layers encompass the digital medium (hardware and content generation) and perceptual sensing, while the top layer performs design translation by converting measured responses into environment adjustments. From this model, we define a Digital Landscape Experience Score (DLES) system with four indices: Dynamicity, Coherence, Experience, and Sensory Perception. These dimensions capture the temporal flow of stimuli, cross-modal consistency, affective engagement, and overall intensity of sensory impressions. In practice, participants interact with VRbased landscape scenarios while multimodal physiological (EEG, eye-tracking, EDA) and behavioral data are collected in real time. A closed-loop pipeline analyzes these data to adjust the virtual environment dynamically, iteratively optimizing the design to match the user's perceptual state. To validate this framework, we conducted user studies and analyzed the results using partial least squares structural equation modeling (PLS-SEM) and linear mixed-effects models (LMM). These analyses test the construct validity of DLES and measure how each index contributes to overall user satisfaction.Implementation of the model and scoring tool confirmed the concept's feasibility. Statistical analysis showed that the four DLES dimensions form a coherent construct: PLS-SEM indicated that each index significantly loads onto a common user-experience factor, and LMM results demonstrated the model's robustness across participants. In expert-evaluated cases, DLES yielded interpretable, actionable insights. For example, in a smart museum immersive exhibition (AR/VR headsets with spatial audio, dynamic lighting, and interactive haptics), the Dynamicity and Experience scores were high (D=0.85, L=0.78, E=0.90, S=0.82), indicating rich stimulation and strong engagement. In a multimedia educational space, experts identified Coherence as the limiting factor; by enhancing visual–tactile synchronization, the coherence index rose from ~0.65 to ~0.80, as reflected in subsequent DLES analysis. In a VR meditation therapy scenario combining nature sounds, subtle haptics, and calming visuals, iterative Delphi scoring converged to DLES values (D=0.65, L=0.88, E=0.92, S=0.80), reflecting very high comfort and guiding designers to balance the sensory elements. Across these applications, experts reported that the DLES model captures core experiential indicators and fosters consensus: iterative Delphi rounds produced convergent, quantitative scores that guided design decisions.This study presents a comprehensive, data-driven framework for multisensory digital landscape design. The four-layer sensoriometric model and DLES scoring system establish a rigorous methodology for quantifying complex sensory experiences in virtual environments. By embedding real-time physiological monitoring and a closed-loop feedback mechanism, the framework enables designers to adapt virtual exhibits on the fly, optimizing dynamicity, coherence, and overall immersion. Empirical validation via PLS-SEM and LMM confirms the framework's structural validity and practical value. Framed in communication-theoretic terms, our model treats design as guiding users' perceptual reconfiguration of the sensory "flux". By delivering explicit, multidimensional feedback, the framework helps translate subjective experience into quantifiable design criteria. By converting complex perceptions into explicit metrics, our approach allows systematic comparison of alternative designs, supporting evidence-based decision-making. Application examples in immersive museum exhibits, VR education, and therapeutic virtual environments demonstrate broad relevance. In summary, this work provides landscape architects with quantitative tools to optimize interactive, multisensory experiences, thereby bridging theory and practice in digital media design.