Skip to main content
In this guide, we explore Amazon Rekognition’s powerful image analysis capabilities. Learn how to leverage label detection, image property analysis, image moderation, facial analysis, and face comparison to enrich your applications with advanced image recognition.

​
Label Detection

Amazon Rekognition’s label detection feature automatically identifies objects, scenes, and activities in an image. By uploading an image, the service detects and labels various objects such as cars, people, skateboards, and shoes. Below is an illustration of a real-world scenario where the service accurately tags objects in an image:
A person is skateboarding on a city street lined with parked cars and trees. The image is part of an Amazon Rekognition demo for label detection.
For instance, by testing with an image of a crosswalk, Rekognition identifies multiple details including road tarmac, zebra crossing, bag, handbag, person, car, wheels, and traffic lights. Every detected object is carefully mapped with its respective location, providing detailed information useful for real-time applications.

​
Image Properties

Another impressive feature of Amazon Rekognition is its ability to analyze image properties. When an image is uploaded, the service returns essential details such as dominant colors, foreground and background distinctions, and overall quality metrics.
The image shows a screenshot of the Amazon Rekognition console displaying image properties for a cityscape photo, including dominant colors and image quality metrics.
By understanding these properties, developers can better manage and optimize images for various use cases like automated image enhancements or content organization.

​
Image Moderation

Managing user-generated content is simplified with image moderation. Amazon Rekognition evaluates images to detect any potentially inappropriate or explicit content. This functionality helps ensure that only appropriate images appear on your website or application.
Image moderation is crucial for platforms that rely on user-uploaded content to maintain community standards and prevent inappropriate material.

​
Facial Analysis

Facial analysis is another robust capability of Amazon Rekognition. By uploading a face-containing image, the service detects facial features and attributes such as gender, age range, facial expressions, and additional details. For example, an analysis of an image of a woman might confirm the presence of a face with 99.9% confidence, classify the subject as female with an age range of 18 to 24, and note attributes like a calm expression with a closed mouth and open eyes.
The image shows a screenshot of the Amazon Rekognition service analyzing a photo of a person standing in front of greenery. The analysis includes attributes like gender, age range, and emotional state.
This detailed facial analysis enables more intelligent applications, from personalized marketing to advanced security systems.

​
Face Comparison

Face comparison allows you to verify whether the same individual appears in different images by comparing a reference image with one or more target images. Rekognition calculates a similarity score for each comparison, providing a precise match—such as a 99.9% similarity score confirming the identity in one of the target images.
The image shows an Amazon Rekognition interface for face comparison, displaying a reference face on the left and comparison faces on the right, with similarity results indicating a 99.9% match for one of the faces.
This feature is especially useful in security systems, user authentication, and systems that require identity verification across different datasets.

​
Additional Capabilities

In addition to the features described above, Amazon Rekognition can identify celebrities and detect text within images. For example, if an image contains a mug with text saying “Monday, but keep smiling,” Rekognition can extract and recognize that text automatically. Integrate the Amazon Rekognition SDK into your application to analyze images in real time, whether they are user uploads or other media. This empowers your application to adapt instantly based on the content of the images.
This overview highlights some of the core features of Amazon Rekognition. As you integrate these capabilities, consider exploring more advanced features to fully leverage the potential of visual recognition in your applications.

Watch Video