Using Custom Algorithms

You can integrate custom algorithms to be used by applications built using the ADAS library

You can use hooks to create a library to support your custom algorithms. To integrate your custom algorithm, you do the following steps:

1. Provide your custom algorithm library with implementations for the following functions from the adas_external_algo.h header file:

   probe(adas_fusion_algo_preferences_t *preferences)

   For information about the function arguments to implement, see “adas_fusion_algo_probe_t.”

   open(uint32_t id, adas_fusion_algo_handle_t *return_handle, char *params)

   For information about the function arguments to implement, see “adas_fusion_algo_open_t.”

   process_sensor_buffer(adas_fusion_algo_handle_t algo_handle, const adas_sensor_properties_t *sensor_properties, const adas_sensor_buffer_t * sensor_buffer, const adas_fusion_buffer_t ** return_buffer)

   For information about the function arguments to implement, see “adas_fusion_algo_process_sensor_buffer_t.”

   close(adas_fusion_algo_handle_t algo_handle)

   For information about the function arguments to implement, see “adas_fusion_algo_close_t.”

2. Create an instance called adas_ext_algorithm_defs and assign your camera driver functions using the adas_ext_algorithm_t opaque type, which is a typedef of the “adas_ext_algorithm.”

   adas_ext_algorithm_t adas_ext_algorithm_defs  = {
   .probe = my_probe_function,
   .open = my_open_algo_instance_function,
   .process_sensor_buffer = my_process_buffer_function,
   .close = my_close_function
   };

   Note: In your algorithm, you must use adas_ext_algorithm_defs to declare your variable name for the structure adas_ext_algorithm_t. The ADAS library links the algorithm to integrate it.
3. Configure the algorithm attribute using the id set to ADAS_ALGO_CUSTOM and the path set to the location of your custom algorithm library in the ADAS library configuration file. Optionally, you can use the parameters attribute to pass a string to the library. For more information about the Algorithm attribute in the ADAS library configuration file, see “Algorithm.” For example:

   {
   ...
   ...
   "algorithm": [
           {
               "id": "ADAS_ALGO_CUSTOM",
               "path": "libfinger_count_algo_example.so",
               "parameters": "five"
               "input": [
                   {
                       "id": "forward_camera"
                   }
               ],
               "output": [
                   {
                       "type": "viewer",
                       "instance": 1
                   }
               ]
           }
       ],
   ...
   ...

Examples of using custom algorithms are available with the QNX Platform for ADAS from the QNX Software Center. See the source directory for more information in your host installation at $QNX_BASE/source/adas-opencv-examples-version.zip. After you extract the ZIP file, you'll find the example at extraction_directory/source_package_adas_opencv_algorithms/apps/adas/adas_algo_plugins.