WELCOME TO
Final Project for Computer Vision Course, Department of Systems and Biomedical Engineering, Faculty of Engineering, Cairo University
About
This project applies everything we've learned in the Computer Vision course and is designed to be user-friendly. It allows users to interact with and try various tools, such as image processing, object detection, and many others, independently.
Features
Pixel provides an opportunity to apply a wide range of computer vision tools, including image processing, object detection, image matching, and image thresholding. These tools offer users a hands-on experience, allowing them to practice computer vision techniques on images and see the results firsthand.
Team
All team members who participated in this project, are affiliated with the Faculty of Engineering, Cairo University.
ImageProcessing
Pixel offers a comprehensive suite of image processing and enhancement tools. These include corner detection algorithms, noise simulation with options like Uniform, Gaussian, and Salt & Pepper noise, and noise reduction filters such as Average, Gaussian, and Median. It also features various edge detection filters, including Sobel, Roberts, Prewitt, and Canny. Additionally, Pixel provides detailed image analytics, including histograms and distribution curves, along with normalization, equalization techniques, and local and global thresholding options.
ObjectDetection
Pixel offers the ability to detect objects, whether they are parametric shapes or require semi-supervised methods. It provides two detection techniques: the Hough transform for identifying parametric shapes such as circles, lines, and ellipses, and the Active Contour Model for semi-supervised shape delineation.
ImagesMatching
Pixel provides image matching services using advanced algorithms like Scale-Invariant Feature Transform (SIFT), Sum of Squared Differences (SSD), and Normalized Cross-Correlation (NCC). These algorithms can detect and match similar features across different images, making them valuable for applications such as plagiarism detection, copyright infringement, and image retrieval.
Image Thresholding &Segmentation
Pixel provides a suite of powerful image processing tools, including advanced thresholding and segmentation techniques such as Otsu's thresholding, optimal thresholding, spectral thresholding, local thresholding, region growing, agglomerative clustering, K Means clustering, Mean Shift Clustering, and RGB-LUV Conversion.
Face Detection &Recognition
Pixel offers face detection and recognition technology using Principal Component Analysis (PCA) and Eigen analysis to identify and analyze faces in images.
