This assignment can be carried out as individual or group projects. The maximum number of members in each group is 3. However, we expect more work and better results from a group with more people, and the responsibility of each group member should be clearly indicated in the report.
In the assignment, students are given an OpenCV-based image retrieval program. The package includes an image database of 1000 images, categorized into 10 groups of images, with each group having 100 images. The package also includes 7 test images for you to test your algorithms as you work on the basic requirements. The package also includes a program, implemented with a naïve image matching algorithm to retrieve the best matched image from the image database given an input image. However, the naïve image matching algorithm is not very accurate. Only 2 of the 7 images can be correctly matched such that the program returns correctly matched images from the database. In other words, the given algorithm cannot successfully retrieve matched images for all 7 test images. In this project assignment, you are asked to improve the matching accuracy of the given algorithm by implementing additional matching criteria, and extend it to include additional features.
There are two levels of requirements for the project, basic and advanced, to cater for students of different backgrounds and interests. The basic requirements are designed for all the students to practice some multimedia programming skills. The advanced requirements are for those students who would like to go further to create an application, and are more flexible in terms of what you would like to do. The basic requirements and advanced requirements account for 80% and 30%, respectively, of the grade for this assignment.
2.1 Basic Requirements (80%)
Students are required to finish all of the following items in the basic requirements:
(1) Improve the number of correctly matched images (20%)
Using the original program, only 2 of the 7 test images will find correct matches. The other 5 images will get wrong matches. Modify the program so that at least 5 of the 7 images will find correct matches.
(2) Modify the above program to retrieve similar images (20%)
Given a similarity threshold value, the program will return a list of images with similarity values higher than the given threshold. The program will save these images to a new folder.
(3) Improve on the Precision (20%)
The target of this requirement is to achieve an average of 60% retrieval precision in requirement (2) for the same 5 test images with correct matches in requirement (1). This means that given a test image, the program will return some matched images. Among these returned images, at least 60% of them are correctly matched. (40% precision gets 5% of marks, 60% precision gets 20% of marks, etc.)
(4) Improve on the Recall (20%)
The target of this requirement is to be able to retrieve an average of 60% of the relevant images in the database. Note that the recall percentage is the average percentage of the same 5 test images with correct matches in requirement 1. (40% recall gets 5% of marks, 60% recall gets 20% of marks, etc.)