Community parks in cold regions play an important role in promoting public physical and mental health. Existing studies have mainly focused on warmer regions and lack refined measurements of the healing quality of winter environments in cold-region community parks. Taking Harbin as a case study, this study first constructs a measurement framework for the healing quality of community park environments in cold regions. A total of 28 typical spatial units across four categories of community parks were selected, and correlation analysis was used to identify the key environmental characteristics affecting the healing quality of different spatial types. On this basis, virtual reality experiments combined with physiological and psychological measurements were employed to quantify the appropriate threshold ranges of key environmental characteristics affecting healing quality, and corresponding optimization strategies were proposed. The results show that key environmental characteristics vary significantly across different spatial types, and that the direction and magnitude of the effects of the same environmental characteristic on healing quality also differ across spaces. Each characteristic has a relatively clear optimization range. Through the technical pathway of “feature identification–threshold quantification–strategy translation,” this study provides a quantitative basis and design reference for the winter environmental optimization of community parks and the creation of healthy living environments in cold-region cities.