Commit 7b511783 authored by Boteju W.J.M.'s avatar Boteju W.J.M.

Update README.md

parent 5ce3ff88
Pipeline #217 canceled with stages
# 2020-045 # 2020-045
Dog Behavioural Monitoring System <h1>DOG BEHAVIOURAL MONITORING SYSTEM
\ No newline at end of file 2020 -45</h1>
<h2>Main Objective</h2>
The main objective of this whole study is to develop a system that has the facility to track and monitor the behavioral changes of dogs according to their breed and age and to make sure our pet dog behaves in a way that is suitable for its breed and age.
<h2>Main Research Question</h2>
According to the previous research work, most of them have used only either sensors or computer vision to track animals, especially dogs’ behavioral changes. Therefore, one of the research problems is how sensors and computer vision can be combined to track dogs’ behavioral changes?
Since the products which are currently in the market, do not compare the dog’s behavior changes based on its breed and age to know whether the dog has the appropriate behavior according to its breed and to its age, the proposed system will be addressed that problem too.
In summary, the following are the research problems that will be addressed through the proposed system:
● How can sensors and computer vision be combined to track dogs’ behavioral changes?
● Check whether our pet dog has the appropriate behavior according to its breed and to its age?
● How to use dog breed identification to track unusual behaviors of a specific dog breed?
● How to track the changes in sleeping hours using low-quality surveillance camera video feed?
● How to identify a change in a dog’s walking activity level if it is undergoing a different health condition?
● How to analyze the barking behavior of the dog and identify the reason for the barking at unusual times?
<h2>Individual Research Questions</h2>
**Dog Breed Identification - IT16175426**
The main research problem in the proposed research component, is how to identify the dog breed using computer vision and specify characteristics of the identified dog breed to track unusual behaviors of the dog.
The following research problems will be addressed through the proposed system.
● How to use computer vision to identify the dog breed?
● How unique characteristics of the dog breeds can be used to track unusual behaviors of the dog?
**Sleeping Pattern Recognition -IT17123150**
The main research problem in the sleeping behavior changes recognition research component, is to track sleep pattern changes in the dog using a surveillance camera and notify the dog owner about the unusual changes through the mobile application.
In summary, the following are the research problems that will be addressed through the proposed system:
● How can sensors and computer vision be combined to track dogs’ behavioral changes?
● Check whether our pet dog has the appropriate behavior according to its breed and to its age?
● How to track the changes in sleeping hours using surveillance camera video feed?
**Walking Pattern Recognition -IT17133678 **
The main research problem in the proposed research component, is to track the walking Pattern changes in the dog using an inertial measurement unit (IMU) sensor and notify the dog owner about the changes through the mobile application.
The following research problems will be addressed through the proposed system.
● How IMU sensors can be used to track dogs’ behavioral changes?
● Dog Activity Recognition (Rest, Walk, Run) since there may be confusion between distinguishing running activity from walking activity.
● Analyze whether the pet dog has the appropriate behavior according to its breed and to its age?
● How to track the changes in walking and motion activity levels using an IMU sensor?
**Barking Pattern Recognition -IT17141352**
There are many devices and apps to control barking as mentioned in the Literature Review above. But dog barking pattern identification to different situations hasn't been implemented. Dogs can be barked when the owner is away. But those reasons which caused barking are not acknowledged by the owner. The situations which caused dog barking cannot be identified by the owner every time even the owner is with the dog. It may be due to health conditions or communication between another dog or after seeing a stranger, etc. In this research following problems will be addressed.
CCTV surveillance cameras will be taken as the main component of collecting data. It was chosen because most places are videoing by them for security purposes. Most of the time dogs may stay within their kennels. Rather than filming their behavior from another video component CCTV camera can be easily used.
● Analyzation of barking behaviors
● Identification of situations which caused barking
● Identification of unusual barking behaviors
● How to record barking using CCTV surveillance camera
<h2>Individual Objectives</h2>
**Dog Breed Identification - IT16175426**
The main objective of the proposed research component is to build up a system that has the ability to identify the dog breed and specify unique characteristics of the identified breed as a help to track unusual behaviors of the dog.
● Sub-Objective 1: Predict the dog breed using a CNN model.
● Sub-Objective 2: Store sleeping hours and activity levels of different dog breeds according to their age.
● Sub-Objective 3: Send sleeping hours and activity levels of the identified dog breed to other components for tracking unusual behavior of the dog.
**Sleeping Pattern Recognition -IT17123150**
The main objective of the sleeping behavior changes recognition research component is to make the system capable of tracking and analyzing the sleeping patterns of the pet dog according to its breed and age.
The main objective in the proposed research component will be covered with the assistance of the following specific sub-objectives:
● Sub-Objective 1: Track the total sleeping hours of the dog from surveillance camera video feed
● Sub-Objective 2: Compare the total number of sleeping hours with the average number of sleeping hours of dog based on its breed and age.
● Sub-Objective 3: Notify the dog owner about the unusual changes in the sleeping pattern of the dog.
● Sub-Objective 4: Add the analysis of the sleeping behavior of the dog to the daily report.
**Walking Pattern Recognition -IT17133678 **
The main objective of the proposed research component is to build up a system that has the ability to track and monitor the walking activity levels of dogs according to their breed and age.
The main objective will be covered with the help of the following specific sub-objectives which will be used with the assistance of machine learning Concepts.
● Sub-Objective 1: Dog Activity Recognition (Rest, Walk, Run).
● Sub-Objective 2: Measure walking activity of the dog using an attached IMU Sensor device.
● Sub-Objective 3: Calculate average walking activity of the dog during normal health conditions.
● Sub-Objective 4: Compare walking and motion activity with average walking activity according to the breed and age and analyze the results.
● Sub-Objective 5: Notify if there is a change in walking and motion activity levels.
**Barking Pattern Recognition -IT17141352**
Main objective of this proposed research is to implement a mobile application which facilitates identifying dog barking behaviors and sending notifications to the owner.
● Record dog barking using CCTV surveillance camera
● Analyzing dog barking patterns
● Compare patterns among themselves to identify the situations which caused barking
● Identification of unusual barking behaviors
● Send notifications to the owner of barkings
● Generate a daily report which has all the records of barking
<h2>Group Details</h2>
**Group ID : 2020-045
Supervisor : Dr. Pradeepa Samarasinghe
Co Supervisor: Ms. Lokesha Weerasinghe**
**Member Details :
Name Student ID Email
Boteju W.J.M IT17133678 it17133678@my.sliit.lk
M.D.P Peiris IT17123150 it17123150@my.sliit.lk
Herath H.M.K. S IT16175426 it16175426@my.sliit.lk
A.K.P Erandi Wathsala IT17141352 it17141352@my.sliit.lk**
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment