The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Voi. XXXVII. Part BI. Beijing 2008
surface, profile vs profile, surface (shape) vs shape, to evaluate
the quality of Airborne LiDAR.
2. FACTORS AFFECTING ACCURACY OF LIDAR
DATA
- Time latency (In such a system as a whole, the time
synchronization of the important role, e.g., GPS and
laser finder system synchronization problems. GPS of
measuring the rate of general in dozens of Hz, but the
laser canner of measuring the rate of in 10 to 100 kHz)
The quality of laser scanning has been studied during the last
few years (e.g. Crombaghs et al., 2002, Ahokas et al., 2003). It
has been shown that the terrain height can be typically collected
within 15 cm. But, in fact, there are large numbers of factors
affecting the quality and accuracy obtained, e.g. the surface
material, flight altitude of sensor and platform, GPS/INS and
observation angle, and so on.
2.1 The criteria of positional accuracy
The NSSDA (National Standard for Spatial Data Accuracy)
uses root-mean-square error (RMSE) to estimate positional
accuracy. RMSE is the square root of the average of the set of
squared differences between dataset coordinate values and
coordinate values from an independent source of higher
accuracy for identical points.
RMSE X = sqrt^ix^-x check i) 2 In] (2)
RMSE y = sqrt [£ { ydata i - y check . f I n\ (3)
RMSE z =sqrt\Y j {z datai -z checki ) 2 In] (4)
Where:
x data,i ’ ydataj ’ Z da,a,i are the coordinates of the i th check
point in the dataset
x check,i ’ ycheck,i ’ z check,, are the coordinates of the i th
check point in the independent source of higher accuracy
VI is the number of check points tested
i is an integer ranging from 1 to n
2.2 The errors in LiDAR
Concerning data accuracy, the major error sources in LiDAR
are the following:
- Positioning of the platform (Satellite has a major role
in GPS positioning reliability. It is quantified by
Positional Dilution Of Precision (PDOP). Poor
satellite, in other words, a high PDOP, generates
inaccurate GPS coordinates.)
- Orientation determination from IMU (The error from
the IMU is as systematic differences on strips depend
strongly on the error from the IMU. The IMU is one
of the main causes of horizontal error in scanned data
points, and errors very often increase or decrease with
consistency in a flight’s direction.)
- Offsets between the laser sensor, INS/POS equipment
and an aircraft platform
- Errors in the electro-optical parts of the laser sensor
- Wrong laser and INS/POS data processing
- Integration and interpolation of the INS and GPS data
- Erroneous data from the reference ground GPS base
stations, or, there are no enough base stations
- Data coordinate transformation
3. WORKING FLOW AND METHODOLOGY
As discussed by the LiDAR equation (Equation 1), there is no
redundancy in LiDAR measurements. This is because, due to
the random nature of LiDAR points, one cannot measure the
exact same point in different strips. Therefore, unlike with
photogrammetric data, one cannot use explicit measures to
assess the quality of LiDAR derived positional information.
Therefore, alternative quality assessment methods are necessary
for this type of data.
According to author's data processing experience and other
literature, a working flow of data processing based on
progressive quality control, is put forward as Figure 2.
Relative
Absolute Quality
Figure 2. Working flow of LiDAR data processing
Against every step in above figure, its methodology could be
discussed following.