SCAiLX-ZB redefines Edge AI cameras with optical zooming

Our SCAiLX-ZB cameras, offering Edge AI capabilities in a Videology Zoom Block bring integration of artificial intelligence algorithms and technologies into our autofocus block cameras. The introduction of neural network inferencing to the Videology Zoom Block cameras makes SCAiLX-ZB smart and powerful Edge AI camera. 

This headway in zoom block cameras enables the camera to perform advanced functions such as facial recognition, object tracking, and motion detection. Additionally, embedded imaging systems with Edge AI cameras can improve image and video processing, allowing for better image stabilization and noise reduction. This can enhance the overall performance and capabilities of the camera, making the smart camera module more versatile and user-friendly.

So, why eDGE AI in a zoom block camera?

To better understand why we are bringing Edge AI at the back of a zoom block camera, one should first understand the differences between AI on the edge and AI in the cloud from an imaging world perspective.

AI algorithms and models running at the device level

Edge AI is artificial intelligence that runs on an embedded device - and is not necessarily connected to the internet to have cloud access. The meaning behind Edge AI and Embedded AI is the same; to make an AI algorithm or model run locally on a device that is built into a system, namely to a camera in our case. These devices allow for the real-time processing of video data without the need for an internet connection to run the AI model. 

HOW EDGE AI differs from AI in the cloud

AI in the cloud, on the other hand, refers to the use of AI algorithms and techniques on video data that is streamed to and processed in the cloud. This approach allows for the use of more powerful resources and larger data sets, but it also requires an internet connection and has a higher latency.

Benefits of using edge AI over using AI in the cloud

• Real-time processing: Embedded AI can process video data in real-time, allowing for immediate actions to be taken based on the results. This is especially important for applications where a delay in processing could have serious consequences.

• Low latency: Since the AI algorithms are running locally on the device, there is less latency in the system. This means that the time it takes for the data to be processed and the results to be returned is shorter.

• Offline capabilities: Embedded AI can continue to function even if there is no internet connection, which is helpful in situations where connectivity is limited or unreliable.

• Privacy and security: Embedded AI can keep sensitive data on the device and prevent it from being transmitted over the internet, which can help to protect the privacy of users and ensure that the data is not intercepted or tampered with.

• Cost-effective: Embedded AI does not require the use of cloud resources, which can be expensive. It may also reduce the need for persistent data storage and transmission costs.

• Low-power consumption: Embedded AI can run on low-power devices, making them suitable for use in battery-powered, portable, or mobile applications.

• Dedicated output format: Embedded AI can output the result of the algorithm in any given format; such as video with overlay tags and boxes, text stream only (json or other) without video, text stream and video stream, and others. In the case of AI in the cloud, the video must always be streamed to the cloud for processing.

SCAILX-ZB: a Zoom Block camera with Edge AI

Videology offers a broad family of zoom block cameras from 10x to 55x optical zoom. All our zoom block cameras come with an industrial-grade sensor with low light performance, coupling the optical zoom lens to the sensor.  Built around a metal chassis, Videology Zoom Blocks are an excellent choice for environments with harsh conditions.

Foreseeing the advancements in AI-based solutions in imaging technologies, Videology has taken a bold step in introducing Edge AI to its zoom block cameras. On-camera AI capabilities of SCAiLX technology allow for real-time analysis of images and video without needing to transmit data to the cloud. 

Explore Sample AI Library Models to Elevate Your Project

Our partnership with Scailable makes SCAiLX-ZB instantly “AI-ready” through the integration of Scailable Edge AI middleware on our devices.

Additionally, a small subset of AI library models is available on SCAILX-ZB for demo purposes.

• 

  Face locator: Locate faces and draw bounding boxes around faces.
• Object detector: Locate people, cars, and bikes. Provide bounding boxes and counts.
• Emotion detection: Primary emotion detection for the whole image; make sure to be close to the camera.
• Face blurring: Using the face locator, we can explicitly blur faces in the resulting output.
• QR code scanning: Try to scan a QR code and see the text stored in the image.

Please note the above-stated AI models are specified using Floating points and optimized for CPU, not NPU at the time this blog post is published. SCAiLX-ZB integration with Scailable middleware is in active development. The following releases of the model catalogue will support INT8 acceleration on NPU.

Leveraging Edge AI at your camera with SCAiLX-ZB depends on your specific needs and use case. Take factors into account such as the camera's features, specifications, price, and compatibility with your existing equipment and software. By utilizing an industrial Videology Zoom Block camera that is AI-ready with your application, you will have the flexibility to adapt the camera to new machine learning tasks and models as they become available. Moreover, you can deploy your own AI models to the edge effortlessly. 

Subscribe to our quarterly newsletter and unlock endless possibilities of SCAiLX-ZB in transforming your AI-powered imaging solutions.