This study is a feature perception study of traditional village landscapes, refines the research method of traditional village landscape features using tourists online comment data, combs the landscape perception features of Xijiang Thousand Households Miao Village from tourists perspectives, digs out the managers’ problems in landscape planning, protection, utilisation, and puts forward the corresponding improvement strategies. In this paper, the Octopus data collector was used to collect online evaluation data from nine target platforms, and ROST CM6 was used to analyse the data for word frequency analysis, semantic network analysis, and sentiment analysis. Finally, using Nvivo12, the data was coded and a multidimensional model of landscape perception was constructed: natural landscapes, cultural and social landscapes, perceived landscapes, seasonal landscapes, and services and infrastructures. The results show that in the landscape perception model, the cultural and social landscape modules are the core factors affecting tourists’ psychological perceptions, with most tourists holding positive emotions towards the scenic spots and a few negative emotions mainly focusing on the relatively weak natural landscape, the unsatisfactory road system and charging system, and the over-commercialised development. In response to these problems, this study proposes corresponding planning recommendations. The results of the study reflect the trend of deep integration of culture and tourism in traditional villages in China, respond to the policy guidance of China’s rural revitalisation policy, and provide useful references and insights for the landscape construction and development of the Xijiang Thousand Households Miao Village and even similar tourist-oriented traditional villages.