VehicleDetect

Classifies aerial land imagery data into different types of surfaces: Water, Tree, Low Vegetation, Barren, Impervious-Other, and Impervious-Road. The classified data is returned as a new raster layer.

This function is part of the machine learning toolset and requires large additional binary files for its operation. If these binaries aren’t detected in the SDK’s working directory at runtime, Global Mapper will attempt to automatically download them. You may receive a pop-up prompt to permit the download.

Vehicle detection can benefit from hardware acceleration if you have a CUDA compatible GPU and choose to enable the aUseCUDADevice boolean parameter.

Parameters:

• aLayerList ([GM_LayerHandle_t32]) – List of imagery layers to use for vehicle detection (if empty, all loaded layers will be used)

• aWorldBounds (GM_Rectangle_t) – Bounds to run vehicle detection within (if None, whole extent of input layers will be used)

• aImageryType (GM_ImageryType_t8) – Type of imagery being used for vehicle detection, classified by its source: aerial or satellite

• aUseCUDADevice (boolean) – If true, will attempt to run on CUDA GPU if available, but will fall back to CPU if not. If false, will run on CPU

• aCustomModelName (string) – (Optional) Filename of a custom model created by previous experimentation to use for this inference. Empty string will use built-in model

• aParentWnd (HWND) – Handle to parent window to attach the inference progress dialog to

• aReserved (uint32) – Reserved for future use; defaults to 0

Returns:

Error Code

Return type:

GM_Error_t32

Returns:

Handle to a new vector layer containing axis-aligned bounding boxes around detected vehicles

Return type:

GM_LayerHandle_t32