辅导 CDS540程序、讲解 Python/Java编程设计
CDS540 Assignment 1 - Text Detection from Images Using OpenCV
Task objective
The objective of this assignment is to use Google Colab, Jupyter Notebook, or any other suitable tool to detect text from images using OpenCV. This will help you understand the concepts of fundamental Computer Vision algorithms and techniques, evaluate the efficiency and effectiveness of computer vision systems, and apply and implement these tools and techniques. Task description:
1.Setup Environment:
oUse Google Colab, Jupyter Notebook, or any other platform you are comfortable with.
oEnsure that OpenCV and other necessary libraries are installed.
2.Load an Image:
oLoad a sample image containing text. You can use images from various sources like scanned documents, street signs, or any image with clear text.
3.Pre-process the Image:
oConvert the image to grayscale.
oApply necessary image processing techniques like thresholding, blurring, or edge detection to enhance text visibility.
4.Text Detection:
oUse OpenCV's text detection methods such as the EAST text detector, or any other OpenCV-compatible methods.
oDraw bounding boxes around detected text areas.
5.Extract and Display Text:
oUse OCR (Optical Character Recognition) tools like Tesseract to extract text from the detected regions.
oDisplay the extracted text and the annotated image with bounding boxes.
6.Efficiency and Effectiveness Evaluation:
oMeasure the performance of your text detection system.
oDiscuss the efficiency (speed) and effectiveness (accuracy) of your approach.
oSuggest potential improvements.
Rubric
Criteria Points
Understand the concepts of fundamental Computer Vision algorithms and techniques 10%
- Demonstrates a clear understanding of image pre-processing techniques
- Explains the chosen text detection method and its working
Evaluate the efficiency and effectiveness of computer vision systems 10%
- Measures and reports the speed of the text detection system
- Evaluates the accuracy of detected text
- Provides a discussion on the system's strengths and weaknesses
Apply and implement the computer vision tools and techniques 20%
- Successfully loads and processes the image
- Implements text detection and draws bounding boxes correctly
- Uses OCR to extract and display text from images
- Code is well-organized, commented, and follows best practices
Total Points 40%
Deliverables
1.A Jupyter Notebook (or equivalent) with your code and in-line explanations.
2.Images used for testing.
3.A brief report (1000-2000 words) evaluating the performance of your text detection system.
Submission Details:
1.You must ensure that all your project files used for this task and the report sit in a directory called “Assignment 1 – Your Name”.
2.All files are required to be uploaded and a link to the “Assignment 1” directory submitted to Moodle.
3.Please make sure that unit Instructor and TA have access to the folder.
4.A link to the demo video of your app running must be submitted.
5.It would be great if you could submit your GitHub link.
6.This is an individual assignment, and you should submit it by 8 pm, Friday, Week 7.
Reference
How to Record the Screen on Your Windows PC or Mac?
How to upload videos on YouTube
Sample video