Remote control of mechanical rat traps based on vibration and audio sensors
(2019) BMEM05 20191Department of Biomedical Engineering
- Abstract
- The main aim with this master’s thesis was to study the possibilities of detecting an event (rat trap going off) by analyzing audio streams from a MEMS (micro-electromechanical systems) microphone or movement patterns from a MEMS accelerometer.
To achieve this goal, a small, power efficient and cost efficient embedded system was built with the help of a development board, a MEMS accelerometer, a MEMS microphone and a metal box where the equipment and the rat traps were placed and positioned.
Data was collected and recorded from both sensors in form of a training set and a test set. The training set was analyzed and algorithms were implemented and optimized in MATLAB for three different cases (Case 1. Microphone, Case 2. Accelerometer... (More) - The main aim with this master’s thesis was to study the possibilities of detecting an event (rat trap going off) by analyzing audio streams from a MEMS (micro-electromechanical systems) microphone or movement patterns from a MEMS accelerometer.
To achieve this goal, a small, power efficient and cost efficient embedded system was built with the help of a development board, a MEMS accelerometer, a MEMS microphone and a metal box where the equipment and the rat traps were placed and positioned.
Data was collected and recorded from both sensors in form of a training set and a test set. The training set was analyzed and algorithms were implemented and optimized in MATLAB for three different cases (Case 1. Microphone, Case 2. Accelerometer and Case 3. Microphone + Accelerometer). The algorithms were based on two steps. In the first step events were detected based on amplitude threshold values. The detected events were then stored as signal segments of a fix interval to ensure that the complete event was captured. In the second step different features were computed to find some characterization of the signal segments to separate correct detections from false detections but also to identify the position of the activated rat trap or whether the rat trap was empty or not.
Finally, the results show that it is possible to detect events where a rat trap has been activated in a quiet environment and the accelerometer had the best performance based on amplitude detection. (Less) - Popular Abstract
- How to use audio from a microphone and vibrations from an accelerometer to detect activated rat traps
Rat problems in a quiet environment? Look no further. There’s no need for complex algorithms. All that is required is an embedded system and a simple algorithm to detect events corresponding to activated rat traps.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8987252
- author
- Solak, Denis LU and Storani, Amin LU
- supervisor
- organization
- course
- BMEM05 20191
- year
- 2019
- type
- H2 - Master's Degree (Two Years)
- subject
- language
- English
- id
- 8987252
- date added to LUP
- 2019-06-27 14:36:07
- date last changed
- 2019-06-27 14:36:07
@misc{8987252, abstract = {{The main aim with this master’s thesis was to study the possibilities of detecting an event (rat trap going off) by analyzing audio streams from a MEMS (micro-electromechanical systems) microphone or movement patterns from a MEMS accelerometer. To achieve this goal, a small, power efficient and cost efficient embedded system was built with the help of a development board, a MEMS accelerometer, a MEMS microphone and a metal box where the equipment and the rat traps were placed and positioned. Data was collected and recorded from both sensors in form of a training set and a test set. The training set was analyzed and algorithms were implemented and optimized in MATLAB for three different cases (Case 1. Microphone, Case 2. Accelerometer and Case 3. Microphone + Accelerometer). The algorithms were based on two steps. In the first step events were detected based on amplitude threshold values. The detected events were then stored as signal segments of a fix interval to ensure that the complete event was captured. In the second step different features were computed to find some characterization of the signal segments to separate correct detections from false detections but also to identify the position of the activated rat trap or whether the rat trap was empty or not. Finally, the results show that it is possible to detect events where a rat trap has been activated in a quiet environment and the accelerometer had the best performance based on amplitude detection.}}, author = {{Solak, Denis and Storani, Amin}}, language = {{eng}}, note = {{Student Paper}}, title = {{Remote control of mechanical rat traps based on vibration and audio sensors}}, year = {{2019}}, }