Software Requirements Specification
for
Classifying Bird Species based on their Sound
Version <1.0>
Prepared by
Name: Cassandra Dias Roll No: ME05 e-mail: fernandes.cassandra@gmail.com
Guide :
Prof. Terence Johnson
Co-Guide : Prof. Sufola Araujo
Course: Date:
M.E.(I.T.) Internet Technology 10th September 2011
SRS for Classifying Bird Species based on their Sound
Contents
1 INTRODUCTION......................................................................................................................................................3 1.1 DOCUMENT PURPOSE ................................................................................................................................................4 1.2 PRODUCT SCOPE.......................................................................................................................................................4 1.3 INTENDED AUDIENCE AND DOCUMENT OVERVIEW.........................................................................................................4 1.4 DEFINITIONS, ACRONYMS AND ABBREVIATIONS.............................................................................................................4 1.5 DOCUMENT CONVENTIONS.........................................................................................................................................4 1.6 REFERENCES AND ACKNOWLEDGMENTS........................................................................................................................4 2 OVERALL DESCRIPTION.....................................................................................................................................4 2.1 PRODUCT PERSPECTIVE..............................................................................................................................................4 2.2 PRODUCT FUNCTIONALITY .........................................................................................................................................5 2.3 OPERATING ENVIRONMENT.........................................................................................................................................5 2.4 DESIGN AND IMPLEMENTATION CONSTRAINTS...............................................................................................................5 3 SPECIFIC REQUIREMENTS.................................................................................................................................6 3.1 EXTERNAL INTERFACE REQUIREMENTS.........................................................................................................................6 3.2 BEHAVIOUR REQUIREMENTS.......................................................................................................................................7
Revisions
Version Author Description of Version Date Completed
1.0
Cassandra Dias
Base version.
09/09/2011
SRS for Classifying Bird Species based on their Sound
1 Introduction
This project attempts to determine the audio features that most accurately classify a given sound sample as belonging to the species Emberiza Schoeniclus. The purpose of this section is to establish the conventions for the rest of this document.
SRS for Classifying Bird Species based on their Sound
1.1 Document Purpose
The purpose of this document is to outline the Software Requirements for a bird species classifier.
1.2 Product Scope
The purpose of the classifier is to enable automatic recognition of the bird species Emberiza Schoeniclus.
1.3 Intended Audience and Document Overview
The document is intended for the professors and for the next phase of the software development life cycle.
1.4 Definitions, Acronyms and Abbreviations
1) SRS: Software Requirement Specification
1.5 Document Conventions
Main Section Titles - Font: Arial Face: Bold Size: 18 Sub Section Titles - Font: Arial Face: Bold Size: 14 Text - Font: Arial Size: 12
1.6 References and Acknowledgments
1) Data Mining Applied to Acoustic Bird Species Recognition by E. Vilches, I.
Escobar, E. Vallejo, and C. Taylor.
2 Overall Description
2.1 Product Perspective
This software is a new self-contained product. It should classify any recording of a birds call as either belonging the species Emberiza Schoeniclus or not. The recording should be a .wav file. It will be most useful for environment monitoring and bird-lovers.
SRS for Classifying Bird Species based on their Sound Pre-process audio samples User Input Feature Selection Build a Classifier
.wav recording of a birds call
Test Classifier
Classifier Output - Belongs to the class or not
2.2 Product Functionality
Input the numerical attributes of the training data Build a decision tree based on the attributes of the training data Input the numerical attributes of the test sample Classify the test sample as either belonging to the species or not
2.3 Operating Environment
Hardware Platform Intel(R) Core(TM)2 Duo E8400 @ 3.00GHz with RAM 1.00GB Operating System Windows (32-bit) Software Applications SoundRuler 0.9.6.0
2.4 Design and Implementation Constraints
The software depends on the audio files being separately pre-processed by the SoundRuler tool. The SoundRuler is an application that retrieves the attributes of the sound file.
SRS for Classifying Bird Species based on their Sound
3 Specific Requirements
3.1 External Interface Requirements
3.1.1 User Interfaces
Classifying Bird Species based on their Sound
SRS for Classifying Bird Species based on their Sound Enter the training data: Build Classifier Brows e
Classifying Bird Species based on their Sound Enter the test sample: Classify Brows e
Class is:
3.2 Behaviour Requirements
3.2.1 Use Case View
Input the numerical attributes of the training data
Element
Actor
Details
User
SRS for Classifying Bird Species based on their Sound
Input the numerical attributes of the training data
Element
Trigger Pre Conditions
Details
1. The classifier is to be built. 1. The numerical attributes of the audio samples are collected. 2. The attributes to be used are selected 3. The numerical attributes and the class are in the format required by the program. 1. The file to be loaded is selected by the program. 1. The file is too large to be read 2. Missing values or file not in proper format. 1. The file is read by the program
Normal course Exceptions Post Conditions
Build a decision tree based on the attributes of the training data
Element
Actor Trigger Pre Conditions Normal course Exceptions Post Conditions
Details
User 1. The training data is ready to be read by the program 1. The training data should be available to the program 1. Attribute selection measures are used. 1. The classifier is not constructed. 1. The classifier is constructed.
Input the numerical attributes of the test sample
Element
Actor Trigger Pre Conditions Normal course Exceptions
Details
User 1. A test sample needs classification. 1. The classifier is ready 1. Attributes are read by the program 1. Missing attributes
SRS for Classifying Bird Species based on their Sound
Input the numerical attributes of the test sample
Element
Post Conditions
Details
1. The test sample is read by the program
Classify the test sample as either belonging to the species or not
Element
Actor Trigger Pre Conditions Normal course Exceptions Post Conditions
Details
User 1. The test sample is read by the program. 1. The classifier is ready. 2. The test sample is read by the program. 1. Attribute selection measures are used. 1. Classifier cannot be accessed. 1. The class is determined by the program. 2. The class is displayed.
