This program can detect objects on the video and print object's names and time inside the video in different formats. Can be useful for creating a YouTube index (Table of Content) or if you have 4 hours video and would like to find a place in this video with the cat. The program generates a list of all unique labels at the end, which can be used as YouTube tags.
Video processing is done locally and your video is not uploaded anywhere. After this page is fully loaded, you can disconnect your PC from the Internet and process your video.
A demo of "how it works" is available by this link and how it can be used is shown in the description of the test video sample here. Test video can be downloaded from here.
The program is written with ML5JS Javascript Machine Learning library.
Object detection results have been copied to the clipboard.
THANK you for your work on image procesing. i had a lot of trouble with tesseraact, and your work helped me to avoid code crashes. haev you ever filtered the camera image?
ReplyDeleteThank you for your comment. What do you mean exactly by "filtered the camera image"?
DeleteI am publishing now another post with motion detection and object detection frpom camera. Maybe it will help.