| |
Record |
Links |
|
Author  |
Vaisanen, T.; Heikinheimo, V.; Hiippala, T.; Toivonen, T., |
|
| |
Title |
Exploring human-nature interactions in national parks with social media photo-graphs and computer vision |
Type |
|
| |
Year |
2021 |
Publication |
The 10th MMV Conference: Managing outdoor recreation experiences in the Anthropocene – Resources, markets, innovations |
Abbreviated Journal |
|
|
| |
Volume |
MINA fagrapport |
Issue |
|
Pages |
248-249 |
|
| |
Keywords |
MMV10 |
|
| |
Abstract |
Understanding the activities and preferences of visitors is crucial for managing protected areas and planning conservation strategies. User-generated geographic information such as photographs shared on social media have emerged as new data sources to complement more traditional visitor information such as on-site surveys. However, analyzing large volumes of photographs manually is a laborious task. Automated analysis of the rich textual and visual content on social media data offers new opportunities for understanding human presence and activities in nature (Toivonen et al. 2019). Approaches for textual and content analysis have been widely developed under the umbrella of conservation culturomics (Ladle et al. 2016). They have been recognized as a useful data source for nature conservation. At the same time, automated analysis visual content has remained rather underexplored when mapping human activities in nature. In this presentation we present our findings of using computer vision methods to explore human-nature interactions from social media photographs and their applicability to visitor monitoring of protected areas. Our main questions are: What types of information can off-the-shelf computer vision methods extract from social media photographs, in terms of activities and preferences of people? Do different visitor groups share different types of photographs from national parks? How does photographic content vary between different types of national parks? |
|
| |
Call Number |
|
Serial |
4311 |
|
| Permanent link to this record |
