WEEK 12: FYP WORKSHOP & MODEL TRAINING AND FINISHING MY PROJECT

Date: 29 - 31 May/1 - 2 June 2023

Content:
During week 12, I participated in an online FYP workshop conducted through Microsoft Teams, focusing on claim and documentation for the Final Year Project (FYP). The workshop provided valuable guidance on documenting the claim, progress report and the process of uploading related documents. This workshop equipped me with the necessary knowledge to ensure proper documentation of my FYP.

Simultaneously, I dedicated time to model training using the YOLOv5 framework. Although initially achieving satisfactory results with precision and recall above 70%, I encountered an issue where the model misidentified the background as fish. To overcome this challenge, I transitioned to YOLOv8, an upgraded version of the YOLO model. The training results with YOLOv8 surpassed expectations, getting precision and recall rates exceeding 90%. With the highly accurate model, I proceeded to develop a Fish Measurement System using Computer Vision techniques, allowing for precise estimation of fish length. Additionally, I utilized the Length-Weight Relationship to estimate fish weight. The Fish Measurement System showcased remarkable results. I do some preliminary experiments and its demonstrated promising outcomes.

Furthermore, in preparation for the upcoming presentation day, I focused on creating a poster that effectively summarized the project. The poster served as a visual representation of the project and stated about project's objectives, methodology, results and conclusions.

Conclusion:
In conclusion, week 12 marked significant progress in various aspects of my Final Year Project. The workshop on claim and documentation ensured proper documentation practices, while the transition from YOLOv5 to YOLOv8 improved model accuracy for object detection. The development of the Fish Measurement System shows impressive results, validating the effectiveness of the Computer Vision techniques.

Appendix:

Figure 1: Model Training Result on YOLOv8

Figure 2: Fish Measurement System Homepage

Figure 3: Fish Measurement System's Measurement Result

Figure 4: Fish Measurement System's Results




Comments

Post a Comment

Popular posts from this blog

WEEK 3: IMAGE BASICS AND OPENCV

Image
Date: 20 - 24 March 2023 Content: In week 3, I continued my learning on image basics with OpenCV, I learned about the basic image operations such as reading, displaying, and saving images using OpenCV. I also practiced working with basic shapes like lines, rectangles, and polygons in OpenCV. I also got to know more about the process of adding text to images using OpenCV, including specifying font types, size, and colours. I learned how to interactively draw shapes on images using mouse and learned event handling functions in OpenCV. Conclusion: In conclusion, in this week 3, I was dedicated to deepening my understanding of image basics with OpenCV. I learned how to perform essential operations on images, such as drawing shapes, adding text, and implementing interactive drawing functionalities using the mouse. Appendix: Below are some output examples of what I have learned. Figure: Output
Read more

WEEK 6: LEARNING OBJECT DETECTION FOR MY SYSTEM

Image
Date: 10 -14 April 2023 Content: In week 6, I was dedicated to expanding my knowledge in the field of object detection using OpenCV and Python. I started by learning the basics of object detection, understanding its significance and practical applications to My Final Year Project named Fish Measurement System. I explored the concept of template matching, a technique used to locate a template image within a larger image, corner detection algorithms, which identify and locate corners or keypoint in an image, detection methods, including popular algorithms like Canny edge detection, to detect and highlight edges in an image. I also learned about grid detection techniques, which involve identifying grids or patterns in an image. I learned about contour detection, it is a process that identifies and extracts the contours or boundaries of objects in an image. Additionally, I tried to make cat detection by using Haarcascade method, I also explored the principles of Haarcascade classifiers an...
Read more

WEEK 2: FYP WORKSHOP & COMPUTER VISION INTRODUCTION

Image
Date: 13 - 17 March 2023 Content: On this week, I attended a FYP Workshop held at TTL2 Hall on UniKL BMI Campus. The workshop focused on providing guidance and clarification on the requirements, expectations, and deliverables of Final Year Project (FYP). The workshop also briefed on the different types of assessment for the FYP and their respective marks weightage. It also stated about the FYP rules and regulations, including guidelines for documentation, submission deadlines, and ethical considerations.  In this week also, I acquired fundamental knowledge about the introduction to computer vision (CV) on image basics and its application, I explored the concept of Numpy and its usage for working with arrays in Python. I also learned how to manipulate and process images using Numpy arrays and developed an understanding of the relationship between images and Numpy arrays and their importance in CV tasks. Conclusion: Overall, this week was focused on attending a workshop that provide...
Read more