The following steps provide a high-level overview of deploying this Ready Solution:
- Place the camera—Depending on the modules being used, camera placement is critical. For the Facial Recognition and the People Counting and Demographics modules, place cameras in a position to capture several frontal views of each person walking towards the camera.
For more information about camera placement, such as the required angles and distance from the object, contact a Dell solution specialist.
- Deploy the EGX software stack on the Dell PowerEdge server—For information about how to deploy the EGX software, see the EGX Stack v1.2 Install Guide for Ubuntu Server x86-64 on GitHub. The steps include installing and configuring Ubuntu OS, and installing Docker, Kubernetes, Helm, and NVIDIA GPU Operator.
- Deploy Deep Vision AI—Deep Vision AI is available for installation either as a Helm chart or by using Docker Compose. Helm charts are required for installation on EGX software stack. If you use Docker Compose, install the NVIDIA Toolkit on the host operating system. In both scenarios, you can download and deploy the Deep Vision AI Admin Console and required software modules as a Docker container.
Note: At the time of publication, not all Deep Vision AI modules are available for deployment through Helm on EGX. In this case, deploy Deep Vision AI modules as a Docker container and through the Docker Compose tool. Contact your Dell sales representative for more information.
- Configure the software modules—When deployed, the Deep Vision AI platform is configured to stream the cameras. For each camera, activate one or more modules. A license key, which can be obtained through a Dell sales representative, is required to activate each camera module.
- Configure cameras and alerts—Finally, add cameras and enable the required software modules in the Deep Vision AI admin console. Configure alerts depending on the use case.