The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B3b. Beijing 2008 
. 16 - 
14 - 
Statistic value of Lt 
Figure 7. Histogram of Monte Carlo simulation process result 
In figure 7, it can give out the critical value for single testing. 
When multiple testing is carried out, Bonferroni adjustment 
needs taking into consideration. According to the criteria 
introduced in section 3.2, critical values at confidence 
level Ct — 0.05 can be easily calculated and the results are 
shown in table 2. 
Times of multiple 
testing 
Critical value 
10 
6.0539 
20 
6.1526 
30 
6.2009 
40 
6.2074 
50 
6.2138 
Table2 Critical value considering Bonferroni adjustment 
From table 2, it is obvious that if more tests are carried out, a 
higher critical value is needed to avoid conservative estimation 
of L f . 
5.4 Node detection result 
After 38 times’ test with a critical value of 6.2061, which is 
calculated from table 2 with a linear interpolation method, 38 
significant local clusters are found. These points are plotted 
along with all FCD points in figure 8. 
In figure 8, all local clusters are found. Besides road 
intersections, some clusters are located along the road, in which 
case it can be assumed that there must be some traffic patterns 
there. This sort of traffic patterns should be paid more attention 
and they should be regarded as nodes of road segments to 
comply with FCD. 
5.5 Determination of final spatial road network 
Based on the strategy discussed in section 4, a pre-processed 
imagery of the study area is needed. Due to the aim of this 
paper is to introduce FCD to road extraction, here the road 
frame is roughly described by hand, which is used to explain 
the candidate road segment selection procedure. Figure 9 shows 
the pre-processed imagery and the final spatial road network 
overlaid with high spatial resolution imagery. 
For convenience, in figure 9(a), the pre-processed imagery is 
given as a binary image, in which white area is the road area 
roughly. In figure 9(b), when the spatial road network is 
overlaid with the high resolution imagery, they match each 
other very well. Nodes of all road segments are highlighted. 
Each node represents a traffic pattern, such as road intersection, 
traffic jam, etc.