33 9 . 7 T i m e After all individual modules were tested for their performance; another test was conducted to
check the time taken by each module to process one frame. This will allow us to analyse where
most of the time is spent and if there is a possibility to reduce it. The stopwatch functions ‘tic,
toc’, mentioned in section 8.1 were used here again for this analysis. Table 4 shows the time
recorded for each module.
Run Module 1 2 3 Average Motion Detection 1.1
1.2
1.1
1.1
Face Detection 0.9
0.9
0.7
0.8
Loading Database 2.2
2.5
2.3
2.3
Face Recognition 7.2
7.6
7.3
7.3
Action Module 0.5
0.3
0.3
0.3
TOTAL 11.9
12.5
11.7
12.0
Table 4 Time taken in seconds for each module Figure 37 Bar plot of time taken by each module Figure 37 shows a plot of the time taken by each module listed in Table 4. This graph provides
us a better understanding of where the majority of the time is spent. The face recognition
module which also includes the database load function takes the majority of the processing
time.
Possible improvements are discussed in section 13 as part of possible future work.
0
1
2
3
4
5
6
7
8
Ti m e tak e n ( s) Average Average
Mukund Agarwal
Face Detection & Recognition System
34 1 0 .