Multi-modal registration
The business goal
Our client is a big pharmaceutical company. Their business goal is to provide a useful tool for their everyday operations on farms in order to improve animal treatment, as well as obtaining the necessary information about animals (weight). In order to accomplish this, they need a tool that could combine some kind of Data Science along with photogrammetry methods.
So, the goal of the project was to develop software for the automatic calculation of animal weight based on the image of the animal with required accuracy no higher than 3%. The calculation of the weight was based on an algorithm of a reconstruction of a 3D model of the animal based on a set of images and estimating the weight using specific animal parameters such as length of the back, size of the torso and amongst other parameters.
Technical side
Our research showed a potential for solving problems within the necessary accuracy limits. Still, this goal was not accomplished due to the necessity of using the correct cameras and the inability to mount these cameras due to the strict pharmaceutical limitations. From the technical side, we used photogrammetry, Computer Vision methods and Deep Learning tools.
Result
As a result, we were able to deliver a Proof-of-Concept that provided good accuracy in most cases (85+%).
Moscow, 117335,
Russia