Virtual reality (VR) can experience a space without creating an actual residential space. Focusing on this point, VR grafting has recently been attempted on emotional research using bio-reactions. However, since studies related to bio-response analysis using VR are focused only on analyzing users" emotions in space, studies on residential space planning supporting mental and physical health are insufficient. Biophilic design provides positive effects such as psychological stability and resilience, so it should be considered when planning a residential space. The method of the study first analyzes the need for biophilic design experience in residential spaces. Second, VR-related research trends and bio-response experimental methods are considered. Third, image analysis algorithms are derived using image collection and deep learning reflecting biofilic experience elements. Fourth, for the VR experience, the image to which the biophilic experience element is applied is analyzed. Finally, we propose the usability of the learning model and the direction of follow-up research by producing a biophilic design-based residential space image. As a result of the study, learning models of "water", "weather", and "plants" were developed, and as a result of image verification, fishing ports, skylights and lighting windows, pots, and trees were recognized. The model of this study is found to be capable of identifying biophilic design experience elements and to be used in the production of biophilic experience-based VR simulations.