The paper provides an overview of a number of crowdsourcing activities that have been spearheaded by researchers at the International Institute for Applied Systems Analysis (IIASA). In particular, we describe the different Geo-Wiki campaigns undertaken as well as an application called FotoQuest Go for in situ field data collection. We then focus on the rapid sorting application Picture Pile and the latest crop type classification approach implemented as part of the Earth Challenge 2020 campaign. Initial results from this latest application of Picture Pile show that a large number of reference data can be collected via crowdsourcing using gamification approaches, which can then be fed into algorithms for crop type mapping. Initial results show an accuracy of around 98% when the crowdsourced data are compared to parcel reference data from the field.