An AI service that analyzes content in images and video.
Extract rich information from images and video
Boost content discoverability, automate text extraction, analyze video in real time, and create products that more people can use by embedding vision capabilities in your apps. Use visual data processing to label content with objects and concepts, extract text, generate image descriptions, moderate content, and understand people’s movement in physical spaces. No machine learning expertise is required.
Text extraction (OCR)
Extract printed and handwritten text from images and documents with mixed languages and writing styles.
Pull from a rich ontology of more than 10,000 concepts and objects to generate value from your visual assets.
Analyze how people move in a space in near-real time.
Run Computer Vision in the cloud or on the edge, in containers.
Easily apply breakthrough computer vision
Add leading-edge computer vision technology to your own apps with a simple API call.
See it in action
Transform your processes
Automatically identify more than 10,000 objects and concepts in your images. Extract printed and handwritten text from multiple image and document types, leveraging support for multiple languages and mixed writing styles. Apply these Computer Vision features to streamline processes, such as robotic process automation and digital asset management.
Maximize the value of your organization’s physical space
Understand how people move in a physical space – whether it’s an office or a store. Create apps that can count people in a room, trace paths, understand dwell times in front of a retail display, and determine wait times in queues. Use these features to build solutions that enable occupancy management, social distancing, optimize in-store and office layouts, as well as accelerate the check-out process. Run the service across multiple cameras and sites.Learn more about this capability
Deploy anywhere, from the cloud to the edge
Run Computer Vision in the cloud or on-premises with containers. Apply it to diverse scenarios, like healthcare record image examination, text extraction of secure documents, or analysis of how people move through a store, where data security and low latency are paramount.Learn about Computer Vision in containers
Build on industry-leading Azure security
Microsoft invests more than USD 1 billion annually on cybersecurity research and development.
We employ more than 3,500 security experts completely dedicated to your data security and privacy.
Azure has more certifications than any other cloud provider. View the comprehensive list.
World-class computer vision at competitive prices
Pay only for what you use with no upfront costs. With Computer Vision, you pay as you go based on number of transactions.
Get started with Computer Vision in 3 steps
"We found Cognitive Services to be the missing piece in the equation, the one that we needed to bring this solution to market and really revolutionize the way people look at video."Katie McCann, Vice President of Product and Engineering, Prism Skylabs
"It didn't take us long to realize Azure Cognitive Services had handed us a powerful set of computer-vision and artificial intelligence tools that we could use to create great apps and new features for our customers in just a few hours."John Fan, Co-founder and CEO, Cardinal Blue Software
Frequently asked questions about Computer Vision
Computer Vision and other Cognitive Services offerings guarantee 99.9-percent availability. No SLA is provided for the Free pricing tier. See SLA details.
Your images and videos are automatically deleted after processing. Microsoft does not train on your data to enhance the underlying models. Video data does not leave your premises and video data is not stored on the Edge gateway where the container runs. Learn more about privacy and terms of usage.
Yes, you can extract one-off images from video content. With "spatial analysis” you can analyze video streams at high-frame rate using cameras connected via Real Time Streaming Protocol.
“Spatial analysis” only detects and locates human presence in video footage and outputs by using a bounding box around a human body. The AI models do not detect faces nor discover the identities or demographics of individuals.
The AI models detect and track movements in the video feed based on algorithms that identify the presence of one or more humans by a body bounding box. For each bounding box movement detected in a zone in the camera field of view, the AI models output event data including: bounding box coordinates of person’s body, event type (e.g. zone entry or exit, directional line crossing), pseudonymous identifier to track bounding box, and detection confidence score. This event data is sent to your own instance of Azure IoT Hub.