COSMONiO® is developing a wide range of image-processing and computer-vision algorithms that operate on any image from the nano scale to galactic scale. Our unique approach ensures that any algorithm works robustly even under the most challenging conditions.
We develop generic image-processing and computer-vision algorithms that work on any type of image. From detecting bacteria to counting stars we follow a universal approach to algorithm design. We believe that since all images are made of pixels, none of them should be treated differently.
COSMONiO’s vision algorithms are:
Low-cost digital microscopes lack autofocus functionality. Industrial products that use these microscopes are often slow since they need to capture every image in the microscope’s focal range before they identify the region of interest.
COSMONiO has developed an ultra-fast and accurate autofocus algorithm that works with any motorised microscope.
Is it possible to automatically identify a specific type of bacteria in a microscope image with multiple focal planes?
COSMONiO has developed an algorithm for bacteria detection and classification on multiple focal planes. Firstly, the algorithm merges multiple partially-focussed images into a single sharp image. Then it identifies and counts specific types of bacteria as illustrated in the image below.