From the perspective of emotional preference, studies have been conducted about landscape healing for subhealth people in the National High-Tech Zone (the study area). Combined with a preliminary practice investigation, Unity 2019 was used to make a digital roaming landscape, and 91 subjects with a history of mental subhealth diseases were randomly invited to participate in the immersive experimentation of the Human–Machine Environment Synchronisation (ErgoLAB) platform. After the experiment, the subjects were invited to fill in a Likert scale as the control group. The interest preference, emotion fluctuation, and healing effect of landscape colour were verified. The results show that: (1) The variation trend of interest reflected the concentration of interest in landscape, and the order of degree of interest, ranging from high to low, went Green > Yellow > Blue > Red > Orange > Purple > Cyan. (2) The subjects’ interest in landscape colour was correlated with the arousal of positive emotions. The correlation between interest in landscape colour and positive emotions, from high to low, went Blue > Green > Yellow, while the correlation between red, cyan, orange, and purple was not significant. (3) The mean skin conductance (SC) fluctuation variance of subjects was 5.594%, which confirmed that the healing effect of digital roaming landscape scenes was significant under the state of low arousal. According to the Likert scale data, subjects’ scores of the healing effect of landscapes, from high to low, went as follows: Green > Yellow > Red > Blue > Cyan > Orange > Purple. The results provide a new method for demonstrating the logical relationship between the digital landscape interest experience–emotional awakening–healing effect and providing a theoretical method and construction scheme for landscape colour configuration in the implementation of landscape healing projects.

In the context of carbon neutrality, it is increasingly important to reduce carbon and increase sinks, and urban green spaces play an important role in carbon sinks. In this paper, we used virtual reality (VR) and photoplethysmographic (PPG) technology to evaluate subject satisfaction regarding urban green space plant community landscape scenes using physiological eye movement and heart rate variability (HRV) data and psychological data obtained according to positive and negative emotional adjectives (PANA). The results of the study showed the following. (1) The physiological data showed the highest visual interest in single-layer grassland. The compound layer of tree-shrub-grass composite woodland communities resulted in the strongest comfort level. (2) The psychological subjective satisfaction evaluation scores were, in descending order: tree-shrub-grass composite woodland (T-S-G) > single-layer grassland (G) > tree-grass composite woodland (T-G) > single-layer woodland (T). (3) The correlation between interest, comfort, and subjective satisfaction was significant, which verified the feasibility of the model of “interest + comfort + subjective evaluation = comprehensive satisfaction”. The results of the study provide theoretical guidance for landscape design based on human perception preferences in the context of carbon neutrality as well as for the implementation of sustainable landscapes to achieve a win–win situation in which carbon sequestration and oxygen release benefits and aesthetics can coexist. The combined physiological and psychological evaluation model can also be applied to other landscapes.