This repository contains notebooks in computer vision tasks using the OpenCV library, including:
Dlib's face detection and shape prediction models are utilized to detect faces in both images and extract facial landmarks. These landmarks are essential for aligning and swapping the faces accurately. After calculating the ConvexHull, the faces are divided into triangles, then the corresponding triangles are resized and swapped.Finally SeamlessClone used to seamlessly integrate the swapped face onto the target image while maintaining natural-looking textures and lighting conditions.
This code provides a comprehensive demonstration of image segmentation and filtering techniques using OpenCV. It explores the use of different color spaces(BGR, HSV, YCRCB and combination of them) and masks to extract Skin areas from images and applies various filtering and morphology operations for image enhancement and noise reduction.
The code will compare each novel image with the reference image and display the results, highlighting potential matches and perspective corrections. It uses computer vision techniques and the SIFT (Scale-Invariant Feature Transform) algorithm to find similarities between images.
This code use various tracking algorithms provided by OpenCV(such as BOOSTING, MEDIANFLOW, MIL, KCF,GOTURN,...). The code is designed to work with video input and utilizes pretrained models for object detection.
The code utilizes a pre-trained face detection, recognition model and matches detected faces with known faces using a similarity threshold.
Other notebooks relating about Pose , Intractive tracking, Connected Components, Equalization, Linear Brightness, ...