We help with the assessment of the technical feasibility of the model based on the customer-specific use case
We could create AI models from scratch or use existing, ready-made solutions.
Our team optimizes the computer vision models for the specific business use cases.
We ensure faster development of the machine learning models, their production, and maintenance.
Our maintenance services include regular model updates as well as retraining to ensure the model is relevant and efficient.
Our CV team provides computer vision data services that include image and video collection, annotation, data augmentation, efficient dataset management for efficient AI model training and deployment.
Optical Character Recognition (OCR)
Extraction of the handwritten, typed, or printed text from the image and its conversion into machine-readable text data.
Military Object Recognition
Detection, identification, and tracking of military-related objects.
Facial Recognition
Identification and verification of a person’s identity based on facial features & biometrics.
3D Reconstruction
Creation of the three-dimensional models using 2D images or video frames.
Barcodes and QR Codes Detection
Detection and decoding of the barcodes & QR codes from video or images.
Video Analytics
Analysis of the video footage or video stream to extract meaningful insights and data (e.g., object count, license plate recognition, traffic monitoring, zone intrusion, license plate recognition (ALPR), etc.).
Anomaly Detection
Analysis and identification of the suspicious events or irregular activities that do not match the standard pattern of behavior.
APIs for OCR:
General APIs for Object, Scene, and Image Analysis
Video Analytics AI
APIs and SDKs for Pose and Gesture Recognition
APIs for Face, Emotion, and Identity Recognition
AI Solutions for Defect Detection and Quality Control as a Part of Industrial Vision
Models for Image Classification and Object Detection
Cloud GPU Servers:
Custom GPU Servers:
On-Premise AI Workstations:
Cloud and Commercial Solutions:
Open-Source and Free Solutions:
Customer: A German leading bank.
Business challenge: Cases have become more frequent when masked attackers with gas cylinders have tried to blow up ATMs to steal money. The customer requested that an additional video analytics feature be added to identify such attempts before they are implemented.
Solution: The computer vision development team developed an algorithm that identified anomalous behavior, appearance (mask or other items covering the face), and the presence of non-standard items (gas cylinders, large packages, etc.) in people heading to ATMs. When identifying suspicious persons, the system sent an alert to the operator, and he accepted this event as normal or called the police to the crime scene.
Project Outcome: Thanks to the use of this solution, the number of ATM robberies by destroying them has decreased.
Customer: European manufacturer of drones and their components.
Business Challenge: The customer was looking for a company with embedded and AI know-how to develop an AI-based drone for autonomous combat missions.
Solution: The Chudovo team was responsible for assembling drone configurations, onboard programming, implementation of an artificial intelligence module, and development of the mobile app for managing mission parameters. The drone system has an intelligent logic that runs live on the drone itself. The drone searches for the targets in the set locations and executes the mission: landing, dropping, or hitting.
Project Outcome: The customer received a ready-made version of the smart drone to launch it into mass production and sales.
Customer: Aldrich by Pleuger Industries
Business Challenge: The main challenge for the customer was the large amount of documentation, namely product specifications and bill of materials (BOM). The company’s main request was to create a solution to digitize its physical archive into an online format.
Solution: The Chudovo team implemented a scalable web-based solution powered by AWS Textract. The solution extracts printed or handwritten text from the source image and generates its online version in a structured way with paragraphs and grouped text, tables with columns and cells with data, checkboxes, etc.
Project Outcome: Fast digitization of the archive and transfer of documents to an online format for their accounting and ease of access if necessary.
Customer: German provider of video surveillance solutions
Business Challenge: The customer company received a request from its client to add an algorithm to the VMS system for simplified automatic counting of the quantity of goods of the corresponding brand passing through a conveyor belt for further distribution to the warehouse.
Solution: The computer vision software development team implemented the video analytics algorithm that conducted the recognition of the brand based on visual signifiers and executed the counting of the arriving items to keep the inventory stock up-to-date.
Project Outcome: The implemented CV solution improved the processes of receiving and accounting for goods arriving at the warehouse in automatic mode.
Automotive
Aerospace and Defense
Smart Cities
