Vision Sensor Solutions
IoT Vision Sensors
High-performance processors, embedded machine learning techniques and low cost CMOS image sensors are enabling a new class of intelligent vision sensors to become technologically feasible and commercially viable. As these technology trends continue expand the world of possibilities for visual intelligence at the edge and IoT devices, the use cases and adoption rate of these is expected to grow exponentially.
In the past, developing an embedded deep learning vision system required the integration of separate camera modules, single board computers and wireless modems or a complex and time-consuming product engineering effort requiring embedded developers with specialized skills in computer vision and machine learning. The DeepViewIoT Vision Sensor Solutions solve these challenges and helps our customers avoid program costs, delays and technical risk by providing a fully integrated, compact, production ready sensor design.
The DeepViewIoT Starter Kits come complete with intuitive development tools, sensor hardware lens and packaging and the configurable embedded firmware required to easily create and evaluate your embedded vision applications, deploy trials and test your real world use case or target market.
When it's time to scale up, the DeepViewIoT Production License allows you to fully customize the entire design including the CV/ML Pipeline, system firmware and hardware design to match your system requirements or we can help you go directly to manufacturing with our Contract Manufacturing partners.
You no longer need to be an expert in AI or Computer Vision to implement a miniature vision sensor capable of identifying and reporting on complex or subjective conditions.
The Sensor development process provides new and experienced users with an end-to-end workflow to create and deploy custom Visual Intelligence functions on the DeepViewIoT Vision Sensors. Examples include: Object Detection, Object Classification, Tracking and Activity/Event Detection and Classification
The DeepViewML Development Suite offers a streamlined approach for creating custom Machine Learning models using your images and datasets. Users are able to curate images or datasets, identify and label key objects of interest for model training and validate performance on the desktop or directly the Vision Sensor. By deploying to the Sensor hardware, you can quickly build and test prototypes in your working environment.
For developers that want to integrate the sensor into a larger system, the DeepView API’s allow you to interface directly with the vision sensor from your equipment or cloud application. Advanced users can also integrate their own firmware application directly into the sensor applications.
Data Set Workspace
Starter Kit | micro edition
Deploying Visual Intelligence at the edge just got a whole lot easier....
24 mm dia.
30mm(w) x 30mm(h) x 38mm(d)
Cellular (NB IoT / LTE-M) communications.
25mm(w) x 35mm(h) x 20mm(d)
The Micro Starter Kit includes a production ready High Dynamic Range (HDR) Vision Sensor powered by NXP's i.MX RT Crossover Series of MCU's and an integrated cellular modem. The kit comes with access to the DeepViewML Development Suite and Demo versions of the DeepView Premium components so you can quickly and easily create a working prototype with your own image datasets.
Once your concepts have been validated for your use case, the full hardware design package is included for volume production as part of a DeepView Premium Subscription. Au-Zone is also available to provide engineering design services to help optimize the Computer Vision & Machine Learning solution, assist with hardware design changes or address any unique product requirements you may have.
Feature Comparison Table