An assessment of the aesthetic value of protected wetlands based on a photo content and its metadata

In this study, we analyze the content and metadata of photos taken in two popular wetlands in South Korea to assess their aesthetic value. Timestamps and geotags are analyzed to determine people's preferred visiting times and locations. Moreover, from the image contents, the photographic subjects are detected using deep learning for object detection, which allows for a quantitative object analysis. The results show that many people took pictures near the starting points of trails in the wetlands. The preferred visiting times depended on the tourism attractions related to specific seasons as well as parts of the day. An analysis of the photos showed that most visitors took pictures of the scenery, such as of the fog and sunsets, a result confirmed by the findings regarding their preferred visiting times, such as morning hours, for the fog, and evenings, for the sunsets. Ironically, the prevalent color of the pictures was dark grayish yellow, which frequently evokes negative emotions such as sadness, disappointment, fear, and contempt, as shown by color-emotion association studies.