Computer Vision

We are experts in providing OpenCV-based scalable solutions on both mobile and desktop platforms under heavy use conditions in critical environments.

Extensive experience with OpenCV implementation of various objects detection, classification and image segmentation tasks using classical machine vision algorithms: SVM and cascade classifiers, optical flow, non-parametric clustering, feature matching and visual words techniques. We also provide OpenCV adaptations of the state-of-the-art deep learning neural-network and autoencoder based models, trained in Google Tensorflow and Caffe frameworks.

Computer vision tasks we successfully handled in recent years include:

● arbitrary objects’ detection and tracking
● 3D scene reconstructions
● ID documents real-time detection, geometric correction and OCR analysis
● face recognition with liveness verification
● image quality enhancement (denoising and superresolution)