One of the most challenging topics of AI has been computer vision techniques. With increasing applications of computer vision witnessed over the last few years, these continue to be used in several new domains, including robotics, surveillance, and healthcare, among others.
In this article, we list down ten popular computer vision projects alongside their available dataset for beginners to try their hands on:-
1| Colour Detection
About: In this project, the goal of the model is to detect every color in an image. This project can be useful in editing pictures and recognizing images. One popular project of color detection is the invisibility cloak using OpenCV.
Dataset: Google-512 dataset.
2| Edge Detection
About: Edge detection is an image processing technique for detecting the edges in images to determine boundaries of objects within images. This technique works by detecting discontinuities in brightness. Some of the common edge detection algorithms include Canny, fuzzy logic methods, etc.
Dataset: The USC-SIPI Image Database
3| Face Detection
About: In this project, the goal of the model is to detect the faces of humans by mapping facial features from a video or an image. There are several steps involved in these projects, such as mapping features, using Principal Component Analysis (PCA), matching the data with the database, and more.
Dataset: IMDB Wiki Dataset
4| Hand Gesture Recognition
About: Hand gesture recognition is one of the critical topics for human-computer interaction. In this project, there are several tasks which are needed to be performed. This includes the hand region, which is to be extracted from the background, followed by segmenting the palms and fingers to detect finger movements. Applications of hand gesture recognition can be in Virtual Reality games, sign languages, among others.
Dataset: Microsoft Kinect and Leap Motion Dataset
5| People Counting
About: The purpose of this project is to count the number of people passing through a specific scene. The applications of this project include civilian surveillance, pedestrian tracking, pedestrian counting, etc.
Dataset: People Counting Dataset (PCDS)
6| Image Segmentation
About: Image segmentation is an essential technology for image processing. This technique can be applied for computer graphics, synthesis of objects, etc. The purpose of this project is to design, implement and test on several regions on a set of images based on the segmentation algorithms.
Dataset: The Berkeley Segmentation Dataset and Benchmark
7| Image Classification
About: The purpose of this project is to classify images where a set of target classes is defined. It is a supervised learning problem where a model is trained to identify the classes using labelled images.
Dataset: The CIFAR-10 dataset
8| Image Colorization
About: Image colorization is a technique that adds style to a photograph or applies a combination of methods to it. One popular project of image colorization is to convert black and white images using OpenCV. The purpose of this project is to produce output colorized images that represent semantics colors and tones by taking an input grayscale image.
Dataset: Image Colorization Dataset
9| Object Tracking
About: The purpose of this project is to develop an object tracking system in a constrained environment. This includes detecting an object from the background and tracking the location of the objects. Object tracking consists of two parts - prediction and correction. The system predicts the object's next state based on its current state, and corrects the state based on the true state.
Dataset: Track Long and Prosper - TLP Dataset
10| Vehicle Counting
About: The purpose of this project is to count vehicles with very good accuracy even in challenging scenarios linked to occlusions and/or presence of shadows. Vehicle counting project can be used for traffic monitoring.
Dataset: Vehicle Image Dataset
Click here to access the list of ten high-quality datasets that one can use for Computer Vision projects.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.