To declare object and event types for your integration, you can either define completely customized types or use predefined types from the taxonomy library. This guide walks you through both methods and shows you how to update your C++ namespaces.

Declare custom object types

Follow these steps to define custom object types to detect people, cats, and dogs by replacing the default Hello, World! object.

1. Open device_agent.h.

2. Rename the symbol kHelloWorldObjectType to kPersonObjectType across your entire project.

   NOTE: If you are using Qt Creator, you can press Ctrl+Shift+R to find and replace across the project.

3. In device_agent.h, set the value of kPersonObjectType to "sample.opencv_object_detection.person".

4. Define constants for the cat and dog objects by adding the following code to device_agent.h:

   const std::string kCatObjectType = "sample.opencv_object_detection.cat";
   const std::string kDogObjectType = "sample.opencv_object_detection.dog";

5. Open device_agent.cpp and update the DeviceAgent::manifestString() method to include the new object types in the objectTypes array:

   std::string DeviceAgent::manifestString() const
   {
       return /*suppress newline*/1 + R"json(
   {
       "typeLibrary" : {
           "eventTypes": [
           {
               "id": ")json" + kNewTrackEventType + R"json(",
               "name": "New track started"
           }
           ],
           "objectTypes": [
           {
                "id": ")json" + kPersonObjectType + R"json(",
                "name": "Person"
           }
           ],
           {
                 "id": ")json" + kCatObjectType + R"json(",
                 "name": "Cat"
          },
          {
                 "id": ")json" + kDogObjectType + R"json(",
                 "name": "Dog"
           }
           ]
       }
   }
   )json";
   }

Declare predefined object types 

Predefined types reside in the default_type_library.json file within the Metadata SDK. To learn more about taxonomy, see metadata_sdk/src/nx/sdk/analytics/taxonomy.md.

To use predefined taxonomy types, whitelist them in the supportedTypes section of your manifest:

1. Open device_agent.h.

2. Rename the symbol kHelloWorldObjectType to kPersonObjectType across your entire project.

3. Define the constants for the predefined types by adding the following code to device_agent.h:

   const std::string kPersonObjectType = "nx.base.Person";
   const std::string kCatObjectType = "nx.base.Cat";
   const std::string kDogObjectType = "nx.base.Dog";

4. Open device_agent.cpp and update the DeviceAgent::manifestString() method with the following changes:

   • Change the objectTypes array name to supportedTypes.

   • Inside the supportedTypes array, change the id keys to objectTypeId.

   • Remove the name attributes, as predefined types include default attributes automatically.

   Your updated manifest method should look like this:

   std::string DeviceAgent::manifestString() const
   {
           return /*suppress newline*/ 1 + R"json(
   {
           "typeLibrary" : {
               "eventTypes": [
                {
                    "id": ")json" + kNewTrackEventType + R"json(",
                    "name": "New track started"
                }
                ]
           },
           "supportedTypes": [
           {
               "objectTypeId": ")json" + kPersonObjectType + R"json("
           },
           {
                "objectTypeId": ")json" + kCatObjectType + R"json("
           },
           {
                "objectTypeId": ")json" + kDogObjectType + R"json("
           }
           ]
   }
   )json";
   }

Update namespaces

Update the C++ namespaces across your integration files to match the new project structure.

Replace the old namespace structure:

namespace nx {
namespace vms_server_plugins {
namespace analytics {
namespace sample {
// ...

} // namespace sample
} // namespace analytics
} // namespace vms_server_plugins
} // namespace nx

With the new namespace structure:

namespace sample {
namespace analytics {
namespace opencv_object_detection {
// ...

} // namespace opencv_object_detection
} // namespace analytics
} // namespace sample

Next steps

After completes these updates, rebuild your integration, copy the new build to the Server, and verify your changes.