📄️ Introduction
Carmen Worker® is a user-friendly, locally hosted version of our cloud APIs. It comes with pre-configured image recognition engines that effectively process most input images out of the box. However, there may be instances where you need to fine-tune the configuration or even switch to a different set of engines altogether. This is where the Carmen® Worker Base Image comes in handy. While providing the same REST API as our cloud service and the standard Carmen Worker®, the Base Image gives you complete control over the details of the recognition process, allowing you to tailor it to your specific needs.
📄️ Creating the Project
First, let's look at the directories and files we need to create:
📄️ Adding Engines
Download the .dat and .so files for the Adaptive Recognition engines you would like to use, and copy them into the dat and lib directories. For example:
📄️ Configuring Engines
Create a gxsd.dat XML configuration file as described by the The Carmen ANPR Engine Reference Manual.
📄️ Creating the Dockerfile
Let's take a look at an example Dockerfile for building a US-specific ANPR service:
📄️ Building the Image
The following bash script makes it easier to build your custom image. Open the build script in a text editor and copy the following code:
📄️ Running the Container
Before running the image, you need to add your runtime configuration to the env file: