Look around for video motion detection

Posted By Oodles Admin | 22-Jun-2016




Curious to know things happening around in your absence? It’s surprising that how the technology has advanced, your presence is not necessary to know about the things happening around. Applications like video motion detection helps you in long run to track the changes in the position.


It is a computer vision technique by which movement can be detected in a video clip or at a specific location. It is noticed by a change in video signal. Oodles technologies is in the race of implementing motion detection. The experts in Oodles Technologies have successfully deployed many Computer Vision Techniques; Motion detection being one of them. Motion Detection is the process of detecting a movement or a positional change in respect to its surrounding.


This feature can be used in many industrial/ scientific applications and mainly for security purpose. It is namely implemented by use of Kurento. Kurento is a WebRTC and a set of API’s which can provide advance media capabilities like Augmented Reality, Speech Analysis, etc.

Media Elements Being the core element of Kurento, establishes recording, loading, transmission and processing of media stream while Media Pipeline establishes the connection and proper flow of media stream from one media stream to another. Kurento’s unique Filter feature helps to achieve the application development.


Understanding the example where Oodelite’s have implemented Motion Detection.


The experts have used an enhanced way to implement Motion Detection via OpenCV. Here the role of filters comes in which turned out to be useful to implement this application. Instead of in-build OpenCV filter usage, a separate OpenCV project was developed, which was then declared and made to call by Kurento.

This application revolved around media recording. Kurento Media Server (KMS) provided the recording capability in a one to many video communication. The recorder Filter element in placed between the WebrtcEndPoints of the Media Pipeline. The Video camera stream, used as a video tag. The video camera stream passes the KMS hence filters which is pushed to OpenCV via http end point, which is further send to Kurento.


Main goal was to develop a opencv project that can be used with WOWZA and other possible streaming server. we know kurento provides 2 kind of filter

  • gstreamer

  • opencv

We didnt built a filter using interface provided by Kurento, as our ambition is to build a motion

detection system for different streaming servers, not limiting to Kurento


How it works?

We have build OpenCv motion detection project using OpenCV. We compare the two consecutive frames from the http stream URL and detect the change in the size. If there is a change that is noticeable with naked eyes above the threshold value, then we come to the conclusion that some motion is detected.

Request for Proposal

Recaptcha is required.

Sending message..