CENG 508
Digital Image Processing
This course offers an analytical exploration of how digital images are acquired, manipulated, and interpreted by computational systems. Starting with the mathematical foundations of image representation and enhancement, the curriculum guides students through the transformation of raw pixels into structured information. By covering essential techniques, from spatial filtering and feature detection to modern deep learning-based classification, this course provides the theoretical and practical framework necessary to solve complex visual problems in engineering and research.
Course Objectives:
- To establish a solid foundation in the computational representation of digital images and basic processing techniques.
- To develop proficiency in extracting robust features and descriptors that allow for reliable image matching and object recognition.
- To introduce classical and modern machine learning frameworks, specifically focusing on the transition from handcrafted features to deep neural networks.
- To explore vision tasks such as segmentation, object recognition, and multi-object detection within static and dynamic scenes.
- To equip students with the skills to process dynamic visual data, focusing on video sequences and object tracking.
Recommended or Required Reading:
Szeliski, R. (2022). Computer vision: algorithms and applications. Springer Nature.
Learning Outcomes:
- Students will be able to implement and evaluate fundamental image processing operations such as noise reduction, contrast stretching, and morphological transformations.
- Students will demonstrate expertise in detecting and matching local features to solve problems related to object category detection.
- Students will gain the ability to design and train deep neural networks to perform automated image classification and multi-class object detection.
- Students will be capable of applying advanced segmentation techniques to partition images into meaningful components.
- Students will master the tools required to analyze temporal data in video streams for motion estimation and object tracking.
| Topics |
| Introduction to image and image understanding |
| Image Representation |
| Basic Image Processing |
| Image Filters and Edges |
| Image Features |
| Keypoint Detection and Matching |
| Image Matching |
| Object Category Detection |
| Object Recognition |
| Review of Machine Learning, Neural Networks, and Deep Learning |
| Image Classification |
| Image Segmentation |
| Deep Object Detection |
| Video Processing and Tracking |
Grading:
- Assignments: 30%
- Project: 10%
- Midterm Exam: 20%
- Final Exam: 40%
Other MS Courses
- CENG 500
- CENG 501
- CENG 502
- CENG 503
- CENG 504
- CENG 505
- CENG 506
- CENG 507
- CENG 509
- CENG 511
- CENG 512
- CENG 513
- CENG 514
- CENG 515
- CENG 516
- CENG 517
- CENG 518
- CENG 521
- CENG 522
- CENG 523
- CENG 524
- CENG 525
- CENG 531
- CENG 532
- CENG 533
- CENG 534
- CENG 541
- CENG 542
- CENG 543
- CENG 544
- CENG 551
- CENG 552
- CENG 555
- CENG 556
- CENG 557
- CENG 561
- CENG 562
- CENG 563
- CENG 564
- CENG 565
- CENG 566
- CENG 567
- CENG 568
- CENG 590
- CENG 608
- CENG 611
- CENG 612
- CENG 613
- CENG 631
- CENG 632
- CENG 641
- CENG 642
- CENG 643
- CENG 651
- CENG 661
- CENG 662
- CENG 663