Astrum Vision solutions will enhance your business

Check our portfolio of computer vision algorithms

As Featured On:

Extract moving parts of a still video

Background segmentation is a separation task, where the goal is to split the image into foreground and background.

Many applications do not need to know everything about the evolution of movement in a video sequence, but only require the information of changes in the scene, because an image’s regions of interest are objects (humans, cars, text etc.) in its foreground. After the stage of image preprocessing (which may include image denoising, post processing like morphology etc.) object localisation is required which may make use of this technique.

2D Localization

Positioning multiple cameras detections on a floor plan

2D localization is planar relationship that transforms points from a camera to 2D plane like floor plan. This camera’s points can be regions of moving objects or detections of selected classes (people, cars, dogs and etc.)

The main advantage of this technique is the versatility that it provides. Cameras should not be mounted in strictly specific and complex locations. We can use our already installed surrvaillance system and enchance it.

Our solution can is not limited in the number of cameras.


Find human skeleton

Human pose estimation is a computer vision-based technology that detects and analyzes human posture. The main component of human pose estimation is the modeling of the human body.

Most used type of human body mode is skeleton-based model. It consists of a set of joints (keypoints) like ankles, knees, shoulders, elbows, wrists, and limb orientations comprising the skeletal structure of a human body. This model is used both in 2D and 3D human pose estimation techniques because of its flexibility.


Find out what human do

This very important and challenging problem to track and understand the behavior of agents through videos taken by various cameras. 

Human activity recognition is formulated as a binary (or multiclass) classification problem of outputting activity class labels. It is a type of time series classification problem where you need data from a camera timeseries to correctly classify the action being performed.

Vision-based activity recognition has found many applications such as human-computer interaction, user interface design, robot learning, and surveillance, among others.

Unprecedented speed

Our algorithms work in real time.

High precision

We provide have state-of-the-art accuracy and recall.

Suitable for edge devices

Our algorithms can run on CPU/GPU and require little computational resources

Custom or pre-trained

You can choose from our pre-trained algorithms or we can build you a custom one.

Flexible pricing

Choose a monthly/annual subscription or pay as you go. Our pricing will fit your business.

24/7 support

Twenty-four-seven support is provided to all of our customers.

Computer Vision Community

A caring computer vision community to help you stay on track

Subscribe now to stay tuned for the latest achievements in computer vision society. If you choose to become part of our community you will regularly receive papers, articles and tutorials.

Sent materials