UrlServer frame recognition

JSON parameters:

{
 "key": "C:\\dpe_001_openvino.annCPU",
 "lprid": "1",
 "features": [
  {
   "left": 1151,
   "top": -2,
   "width": 107,
   "height": 23,
   "image": "....",
 },
..... Any number of frames
  {
   "left": 1145,
   "top": 16,
   "width": 112,
   "height": 25,
   "image": "....",
  }
 ]
}

where

key - an arbitrary unique key. It is necessary to prevent reloading of the neural network file if several LPR channels use the same neural network file and device;
lprid - identifier of the LPR channel in Auto-Intellect (see Setting up the LPR channel);
left, top, width, height - number plate coordinates;
image - vehicle image prepared for Vehicle Type Recognition Module.

Single-frame response example:

{
 "results": {
  "color": {
   "error_code": 200,
   "tags": [
    {
     "id": 0,
     "color": "",
     "probability": 1
    }
   ]
  },
  "mmt": {
   "error_code": 200,
   "tags": [
    {
     "id": 0,
     "make": "",
     "model": "",
     "type": "car",
     "probability": 1
    }
   ]
  }
 }
}

where

type - vehicle type:
- noise - noise;
- bus - bus;
- car - car;
- motorcycle - motorcycle;
- small bus - minibus;
- truck - truck.

Page tree

Documentation for Auto Intellect 5.6. Documentation for other versions of Intellect is available too.