Zapp: Learning about the Distant Landscape

A successful application area of mobile technology for learning has been to provide location-based guides that inform students or tourists about their immediate surroundings. In this paper we extend this location-based learning to the distant landscape, so that a visitor to an unfamiliar area can ask “what can I see over there?”, or can annotate the landscape by taking photos of distant features and adding text or audio notes that are the automatically located as points on a digital map. We describe a system, named Zapp, for learning about the distant landscape. It uses a line of sight algorithm computed over a digital surface model stored on a smartphone to determine which distant feature is showing in the centre of the smartphone camera screen. In ‘query’ mode the system can inform the user about pre-stored elements of a landscape such as names of rock formations. In ‘capture’ mode the user can store a note about a distant feature, linked to a photo and to its map coordinates. A trial of the system with university students has demonstrated its usability and usefulness in interpreting the geology of a rural landscape. BACKGROUND Finding out about one’s immediate environment is a well-recognised application of mobile learning. This can be achieved through the use of a handheld device with location sensing technology, such as GPS or phone cell positioning, running an application to deliver information in text, audio or multiple media about the user’s current location. Examples of such outdoor mobile guides include Cyberguide (Abowd et al., 1996), GUIDE (Cheverst, et al., 2000) and Caerus (Naismith, Sharples & Ting, 2005). See Kray, Baus and Cheverst (2005) for a survey of location-based guides. While there can be problems with accurate positioning, particularly in an urban environment with areas of ‘GPS shadow’, location-based guides have provided a foundation for research and development in contextual mobile learning. Several research prototypes (e.g. Quinn & Cartwright, 2009; Bradley et al., 2010) have been based on the Mscape platform (Stenton et al., 2007) produced by HP Labs in Bristol to develop mobile media guides and games. Another approach is to augment the natural landscape with instructional information and stimuli for inquiry learning. The best-known is Ambient Wood (Rogers et al., 2004) where information and instruments were embedded in a woodland for children to carry out investigations into its biological processes. All these examples assume that the learners are gaining information about their immediate surroundings. But what if they want to learn about a distant object, landmark, or area in the landscape? Some examples of such need include the following: tourists interested in information about visible landmarks such as the names of distant mountain peaks; students on a geology field trip learning about rock formations in the surrounding landscape; visitors to a heritage site exploring how an area looked in historic times; or for general users of camera phones, being able to log one’s photographs of distant landscape features on a map to show both the photographer’s position and the location of the distant feature. Current technologies can offer some assistance by two different methods: augmented reality based on device position, orientation and tilt; and image recognition of landmarks. Augmented reality systems such as Layar (http://www.layar.com/) and Wikitude (http://www.wikitude.com/) show information about nearby buildings or landmarks overlaid on the phone camera screen. The problem is that these applications only show what is in the general direction of the camera, not what the user can actually see. If the user’s line of sight to a target building or landmark is blocked by another building or object, then this can cause confusion or misinformation. The other approach (with Google Goggles being most widely used example, http://www.google.com/mobile/goggles) uses image recognition software to identify a landmark within a photo taken by the phone camera and then search for relevant information about it. For