Sample Narrative for FOCI Poster
The proposed system uses the standard client-server model as shown
in Figure 1. A scientist in the field (1) equipped with a mobile device
(i.e., a smart phone, pda or tablet running on the Google Android or
Apple iOS platforms) takes digital photographic images of arthropods,
weeds, or plant diseases. Software on the mobile device (2) alerts the
scientist if the image is of insufficient quality (blurred, lacking in
contrast, too dim, etc.) and suggests a retake. When a suitable image
is taken, (3) the scientist may append a text message and/or an audio
clip before uploading the information to the web/database server. Such additional information
may provide helpful guidelines to the classification software and facilitate the classification
process. GPS information marking the location of the upload will be determined and sent
automatically.
On the backend (4) all information is stored in the system database and sent to the compute
server (5) for classification. This information includes the image taken supported by any text and
audio information provided by the scientist. The classification algorithms are computationally
intensive requiring one or more high-performance compute servers. Results of the classification
algorithms are (6) sent to the server, saved in the database, and (7) sent to the client in text and
audio (optional) format. Links to sites providing additional information about the arthropod,
weed, or plant disease are optionally provided.
All information collected is stored in the system database and can be available for subsequent
review (8) either on a mobile device or on a laptop or desktop computer. GIS and Google maps
with pushpin markers will pinpoint locations of uploaded information. Map information can be
grouped in various ways: by date and time, by biological classification, by geographical location,
or by other criteria suggested by the participating scientist.