Research Computer Vision library on the web

[sacdallago]

https://trackingjs.com/docs.html#introduction

[sacdallago]

http://wesbos.com/demos/html5-face-detection/

[fayeli]

headtrackr.js: Javascript library for headtracking via webcam. It uses a Viola-Jones type detector from ccv library to detect the faces, then implements Camshift algorithm to track the object based on colour histogram. https://github.com/auduno/headtrackr Details: http://auduno.com/post/25125149521/head-tracking-with-webrtc

[fayeli]

JSfeat: A Javascript library with implementations of many directly relevant image processing or computer vision algorithms, such as feature detection or pre-trained face detectors, which can help finding where the user is on the webcam. It also has geometric transformations (e.g. homography, affine) which we can use for mapping a DICOM image. http://inspirit.github.io/jsfeat/

[fayeli]

clmtrackr: A Javascript library for fitting facial models to faces in videos or images. It requires JSfeat for initial face detection, and provides the fitted facial model which can allow for more accurate mapping than simple overlaying images. https://github.com/auduno/clmtrackr Demo: https://auduno.github.io/clmtrackr/face_mask.html

[fayeli]

node-opencv: NodeJS binding for opencv. Should support all the computer vision features of opencv, such as the Viola-Jones face detector and others relevant functions. https://www.npmjs.com/package/opencv Tutorial: https://www.sitepoint.com/face-detection-nodejs-opencv/

[fayeli]

js-objectdetect: Javascript library implementing Viola-Jones for object detection https://github.com/mtschirs/js-objectdetect

[fayeli]

ccv: Javascript library implementing many computer vision algorithms https://github.com/liuliu/ccv