Lab3.markdown

@@ -202,6 +202,7 @@
 > sound (see video).
 >
 > [![image alt text](http://img.youtube.com/vi/UMSrPEmQrwE/0.jpg)](http://www.youtube.com/watch?v=UMSrPEmQrwE)
+> ##### Fig. 7: Video of the clap detecting robot in action.
 >
 >```
 > while (!Button.ESCAPE.isDown()) {
@@ -229,10 +230,10 @@
 >			Thread.sleep(5);
 >		}
 >```
-> Fig. ??: Inner loop of the ClapCar program.
+> ##### Fig. 8: Inner loop of the ClapCar program.
 >
 > ![G3](http://gitlab.au.dk/uploads/group-22/lego/8b5ba5b544/G3.png)
-> Fig. ??: Graph showing the profile of a single clap of sound level over time.
+> ##### Fig. 9: Graph showing the profile of a single clap of sound level over time.
 >
 > We noticed that it is possible to trick the robot into thinking it hears a clap by shouting a short word to the sound sensor. This is due to the fact that the robot only distinguishes between volume and duration of a sound meaning that different sounds can actually result in the same sound profile. The graph below [fig. ??] shows a series of claps showing three distinct peaks. Each peak is a registered clap. The smaller peaks will not result in a beep as the sound profile does not match that of a clap. 
 >
@@ -247,7 +248,7 @@
 > drive towards the location with the loudest sound level.
 >
 > ![P2](http://gitlab.au.dk/uploads/group-22/lego/ed19ec2872/P2.png)
-> Fig. ??: Picture showing how we mounted the two sound sensors on the NXT robot.
+> ##### Fig. 10: Picture showing how we mounted the two sound sensors on the NXT robot.
 >
 > The two sound sensors were mounted on the opposite sides of the robot facing away from each other at a 90 degree
 > angle (perpendicular to each other), as can be seen in fig. ??.