charger. Professional drones can charge faster (half an hour) than consumer drones.
It doesn’t mean theconsumer drones are toys and not serious enough for professional
use. Although the professional configuration is much more expensive, they might be
equipped with high-tech components which you really don’t need. A company from
China, DJI, has seized the first gold as a drone manufacturer (Atwater 2015; Lee and
Choi 2016). Their products are the best option you can find in current market. The
price varies dramatically depending on model and configuration. Most of them are
appropriate for data collection over agricultural fields. During a fly task, operators
can set up the observation mode, plan a path, let the drone fly along, manipulate the
remote control, and down-stream the real-time images/videos to a tablet. Most
equipped cameras on UAV are installed on a gimbal and can rotate to shoot
images from a perfect angle. The drone has memory onboard, and all the captured
data are automatically stored there. Once the drone lands, operators can take the
memory and copy the data to storage devices. To facilitate the post-fly data
processing, the drone companies and some other image processing companies
developed a number of software (Newman 2013). The processes include image
fusion, mosaicing, georeferencing, calibration (need ground control points), interpolation, etc. NDVI and other vegetation index products can be derived from the
calculation among the optical bands and infrared band. DEM standard products may
also be achieved if the drone camera takes stereo¼pair images. Until now, a whole
set of hardware, software, theories, and techniques have become mature for using
drones in precision agriculture. In the past 5 years, drone images can be seen
almost in every study and use for agriculture. In the future, the use of drones will
be more, and the application of drone images will be everywhere. Some military
fuel-powered drones can already fly up to 36 hours before returning to base. We
believe as the rapid progress on development of low-cost long-duration battery and
long-distance remote control, the observation region of civil drones will be greatly
extended in next few years. Please refer to Boucher (2015) for more progresses on
domesticating the military drones for civil use.
However, the publicity and openness of drone data are not as good as satellite
data. The use of a drone is field specific. Similar to aircraft-based data, the drone data
are limited in both temporal and spatial extents and has neither global nor national
products. It seems that small region is the birthmark of civil drones. It is very difficult
for users to discover and order low-cost data sources on a specific field, especially
when the field has no record of drone flying. Commonly, data consumers either buy a
drone to obtain the data themselves or outsource the project to an experienced
company. The resulted data products are claimed by the initial consumers and
archived in their private storage facilities. Outsiders have no route to access them.
This is the new normal. In contrast, some government programs and non-profit
foundations have done some open UAV tasks and published the sample datasets
online to boost the contribution from citizen scientists or research institutes. But such
datasets are rare, small, incomplete, low-quality, and meaningless for consumers
with an interest in a different field. Lots of work need to be done to form a global
drone image database for searching, ordering, purchasing, processing, and
downloading.
52
Z. Sun et al.
It doesn’t mean theconsumer drones are toys and not serious enough for professional
use. Although the professional configuration is much more expensive, they might be
equipped with high-tech components which you really don’t need. A company from
China, DJI, has seized the first gold as a drone manufacturer (Atwater 2015; Lee and
Choi 2016). Their products are the best option you can find in current market. The
price varies dramatically depending on model and configuration. Most of them are
appropriate for data collection over agricultural fields. During a fly task, operators
can set up the observation mode, plan a path, let the drone fly along, manipulate the
remote control, and down-stream the real-time images/videos to a tablet. Most
equipped cameras on UAV are installed on a gimbal and can rotate to shoot
images from a perfect angle. The drone has memory onboard, and all the captured
data are automatically stored there. Once the drone lands, operators can take the
memory and copy the data to storage devices. To facilitate the post-fly data
processing, the drone companies and some other image processing companies
developed a number of software (Newman 2013). The processes include image
fusion, mosaicing, georeferencing, calibration (need ground control points), interpolation, etc. NDVI and other vegetation index products can be derived from the
calculation among the optical bands and infrared band. DEM standard products may
also be achieved if the drone camera takes stereo¼pair images. Until now, a whole
set of hardware, software, theories, and techniques have become mature for using
drones in precision agriculture. In the past 5 years, drone images can be seen
almost in every study and use for agriculture. In the future, the use of drones will
be more, and the application of drone images will be everywhere. Some military
fuel-powered drones can already fly up to 36 hours before returning to base. We
believe as the rapid progress on development of low-cost long-duration battery and
long-distance remote control, the observation region of civil drones will be greatly
extended in next few years. Please refer to Boucher (2015) for more progresses on
domesticating the military drones for civil use.
However, the publicity and openness of drone data are not as good as satellite
data. The use of a drone is field specific. Similar to aircraft-based data, the drone data
are limited in both temporal and spatial extents and has neither global nor national
products. It seems that small region is the birthmark of civil drones. It is very difficult
for users to discover and order low-cost data sources on a specific field, especially
when the field has no record of drone flying. Commonly, data consumers either buy a
drone to obtain the data themselves or outsource the project to an experienced
company. The resulted data products are claimed by the initial consumers and
archived in their private storage facilities. Outsiders have no route to access them.
This is the new normal. In contrast, some government programs and non-profit
foundations have done some open UAV tasks and published the sample datasets
online to boost the contribution from citizen scientists or research institutes. But such
datasets are rare, small, incomplete, low-quality, and meaningless for consumers
with an interest in a different field. Lots of work need to be done to form a global
drone image database for searching, ordering, purchasing, processing, and
downloading.
52
Z. Sun et al.