Starting with DetectorPack 3.10, a utility was added to the VT LPR add-on (see Installing DetectorPack add-ons), which allows you to create GPU neural network caches without using Axxon One. The presence of the cache speeds up the initialization and optimizes video memory consumption.

To optimize the operation of the License Plate Recognition VT on GPU, do the following:

1. Stop the Server (see Shutting down a Server).
   

   If the system has the software running on GPU, it is necessary to stop its operation.

2. Create the GPU_CACHE_DIR system variable (see Appendix 9. Creating system variable) by specifying in the Variable value field the path to the cache location with an arbitrary folder name. For example, D:\IX_GPU_cache. The specified directory will store the cache for all used detection tools and neural networks.

3. To call the utility, open in the command line: C:\Program Files\Common Files\AxxonSoft\DetectorPack\VitGpuCacheGenerator.exe and press Enter.

4. Specify the required countries in alphabetic form.
   
   After you selected all required countries, enter done.
5. The default resolution is 1920x1080 pixels. If you want to change the resolution, enter y and the required resolution.
   
   To leave the default value, enter n.
   
6. Specify the minimum width of the license plate according to the actual size of the license plate in the frame as a percentage. The minimum allowed value is 0, the maximum value is 100.
   
   
7. Specify the ID of the required GPU (see Selecting Nvidia GPU when configuring detection tools) and press Enter.
   
   

As a result, the process of creating cache starts, which will take about 30 minutes. The duration of the process depends on the selected types of cache, number of countries and NVIDIA GPU model.

Optimizing the operation of the License Plate Recognition VT on GPU is complete.