-
Notifications
You must be signed in to change notification settings - Fork 0
/
notes.txt
33 lines (27 loc) · 1.31 KB
/
notes.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
FYP Notes:
25/2/2023
Scope:
Goals:
Challenging:
- Localise more than one source,
- Incorporate a communication protocol with other robots,
- Detect a quick short sound, e.g. a snap,
Ideal:
- Localise three-dimensionally with range and bearing,
- Help the overall localisation of the robot,
- Have real-time signal-processing and localisation, i.e. processing in parallel with acquisition,
- Have a wide bandwidth, e.g. ultrasonic sounds as well as audible,
- Detect a more varied sound, e.g. voice,
Basic:
- Localise at least two-dimensionally with range and bearing,
- Process and acquire at least serially, i.e. switch between the two,
- Communicate with the NUC,
- Detect at least audible sounds,
- Detect at least a strong constant sound, e.g. a whistle,
Questions:
- What is the method of localisation, either time-difference-of-arrival (TDOA), beamforming, or something else?
- Where is the computation? Should it be on an STM32, DSP, SBC, NUC, etc.?
- What microphones are best? They should preferrably be omnidirectional.
- Which kind of signal is best? Should it be analogue or digital? The sampling requirements will set this criterion.
- What level of electronics design is there? If analogue microphones are to be used, then fitlering may be needed.
4/3/2023