Agriculture Reference
In-Depth Information
that needs to be addressed is data synchronization both temporally and spatially.
Because some of the collectable data do not directly present relevant production
information, that is, not directly observable, it is important to represent the collected
data in a form observable by users, that is, to synchronize the data to show the
relevant information. Two other critical functions in data management are to con-
vert the collectable data into observable information, then to extract the actionable
data from the observable data. The basic requirement for the data management soft-
ware is the capability of extracting actionable data from the collectable data using a
“transparent-to-user” way that provides users operational instructions “on-the-go.”
Data management software of precision agriculture often involves four steps: data
collection, data integration, data mining, and data visualization.
3.4.2 D
ATA
C
OLLECTION
Field data collection is very important for efficient automated precision operations in
agriculture production. Data collection software is a part of a data acquisition system
that typically cooperates with data acquisition hardware to gather data from indi-
vidual or multiple sources. Typical data collection software for field data acquisition
often includes the elements of (1) software drivers that allow a higher-level computer
program to interact with a hardware device; (2) data telecommunication between
sensing and data processing units to make the sensed data useful to the operators;
and (3) data storage, which records or retrieves data for further application.
A software driver simplifies programming by acting as a translator between a
hardware device and the application or operation systems that use it. Programmers
can write the higher-level application code independently of whatever specific hard-
ware device. In most applications, software drivers are often provided by acquisi-
tion hardware suppliers that can usually be briefly integrated into whole software
systems. Because of the nature of the operation, many field data collection software
packages carry both software drivers and data telecommunication elements.
Field data are normally collected either during regular production operations,
such as site-specific yield data collection during harvest, or from special data col-
lection operations, such as soil compaction surveys. Effective and efficient data
telecommunication capabilities are critical for modern agricultural applications.
Numerous field data collection software has been developed for both research and
product development. A few examples of research applications include the software
for a remote temperature monitoring device used for long-term wireless monitoring
of temperature in bulk vegetable stores (Misener et al., 1989); for a 802.11b internet
protocol-based data communication between in-field tractors, and between tractors
and in-office computers (Will et al., 1999); for a Pocket PC (PPC)-based field data
management system to acquire and analyze field information (Fang and He, 2008);
and for mapping a large number of field images using GPS-positioned image col-
lection systems for generating a weed coverage quantification map from 1000 field
images with 5000 GPS coordinates within 30 min (Wiles, 2011).
One challenging factor in field data acquisition for precision/automated agricul-
tural operations is its information incompleteness from one time and/or one source,
in addition to its very poor repeatability. One practical way to solve the problem is to
