Aaeon/Wahtari SmartCam Juno for Computer Vision

Aaeon/Wahtari SmartCam Juno for Computer Vision

One of the use cases of Wahtari's smart camera is Quality Control shown in real-time. You can present an object in front of the camera and the system shows whether there is a problem with it or not (for example a broken object, or color difference, etc.) Wahtari prepared the AI model for this demo in less than 24 hours, using their in-house AI platform. Due to this, the ability to efficiently detect objects on its self-made smart camera can be done with 30 to 120 FPS. Wahtari's smart camera is based on UP Core and UP AI Core X which uses Intel Apollo lake processor and Movidius Video accelerator. Together with Wahtari's end to end AI platform, you are able not only develop Quality Control but also any computer vision tasks ranging from license plate recognition up to analysis of medical images or even facial recognition in the public security sector.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • An all in one solution
    Wahtari provides an all in one solution including AAEON's HW and Wahtari's SW parts which saves a lot of effort when building your own AI application
  • An upgradable solution
    The smart camera's configuration is not fixed. If you want to have higher resolution or more FPS (frame per second) it is possible, as each part of the camera can be swapped due to its modular design.
  • Limitless possibilities for computer vision tasks
    Combined with Wahtari's end to end AI platform: Juno, you are able to develop any computer vision tasks from license plate recognition to analysis of medical images or even facial recognition in the public security sector for example.
  • An easier path to build image recognition applications
    Wahtari only requires some images to train their AI model and build up the AI applications. Wahtari is able to solve all the aspects including the ready to use camera and provide POC in a very short amount of time.



Brazil China (PRC) France Germany India Japan Korea Mexico Other - Asia Pacific Other - Europe and Africa Other - North and South America Russia Taiwan United Kingdom United States Worldwide Intel® Movidius™ On-premise (Private Cloud, Other) Caffe2 MXNet TensorFlow Torch/PyTorch Manufacturing Transportation and Warehousing Models can be trained - data input only required Models can be trained - requires labeled data AI Appliance AI Platform as a Service (AI PaaS) AI Software/SaaS Linux Compute Library for Deep Neural Networks (clDNN) Intel® Math Kernel Library for Deep Neural Networks (Intel MKL-DNN) Intel® Distribution of OpenVINO™ toolkit Faster RCNN MobileNet Proprietary ResNet50 SSD SSD-VGG16 Yolo Anomaly Detection Data Preparations and Management Facial Detection/Recognition/Classification Image/Object Detection/Recognition/Classification Medical imaging, analysis and diagnostics Smart City Video Surveillance and Analytics Deep Learning