Testing Overview¶
This report presents the testing procedures and results for the SignBridge project, a real-time Sri Lankan Sign Language (SLSL) detection and translation web application. The purpose of testing was to evaluate the functionality, reliability, usability, performance, and security of the system. The SignBridge application uses machine learning and computer vision technologies to recognize hand gestures through webcam input and convert them into text in real time. Since the system combines frontend, backend, machine learning, and real-time communication components, comprehensive testing was required to ensure system stability and usability. The testing process involved:
Functional testing
Security testing
performance testing
Accuracy testing
Usability testing
Automated testing
Manual testing
Objectives of Testing¶
The primary objectives of testing were:
To verify that all functional requirements work correctly
To identify system errors and bugs
To validate reliability of the machine learning model
To evaluate real-time system performance
To ensure the system handles invalid inputs securely
To confirm that the application provides an acceptable user experience
Testing Environment¶
Operating system
Windows 10/11
Backend Framework
Flask
Frontend Technologies
HTML, CSS, JavaScript
Machine Learning Framework
TensorFlow
Testing Framework
pytest
Browser Automation
Selenium
Database
SQLite
Hardware
Standard laptop with a webcam
Browsers used
Google Chrome/Brave
Testing Methodologies¶
The testing process was divided into two major categories:
Manual Testing¶
Manual testing was conducted to validate:
Real-time webcam functionality
Gesture recognition
User interface responsiveness
System userbility
Real-world performance
Automated Testing¶
Automated testing was conducted using pytest and Selenium to validate:
Backend functionality
API responses
Authentication features
Route handling
Error handling
User interface behavior