vagheshpatel commited on
Commit
f21564b
·
verified ·
1 Parent(s): 72ef152

Sync loitering-detection from metro-analytics-catalog

Browse files
Files changed (1) hide show
  1. README.md +21 -5
README.md CHANGED
@@ -6,7 +6,7 @@
6
  |---|---|
7
  | **Category** | Object Detection + Tracking + Zone Analytics |
8
  | **Source Framework** | PyTorch (Ultralytics) |
9
- | **Supported Precisions** | FP32, FP16, FP16-INT8 |
10
  | **Inference Engine** | OpenVINO |
11
  | **Hardware** | CPU, GPU, NPU |
12
  | **Detected Class** | `person` (COCO class 0) |
@@ -44,6 +44,10 @@ python3 -m venv .venv --system-site-packages
44
  source .venv/bin/activate
45
  ```
46
 
 
 
 
 
47
  ---
48
 
49
  ## Getting Started
@@ -155,7 +159,7 @@ pipeline_str = (
155
  f"gvametaconvert add-empty-results=true ! queue ! "
156
  f"gvafpscounter ! "
157
  f"gvawatermark ! videoconvert ! video/x-raw,format=I420 ! "
158
- f"x264enc ! h264parse ! "
159
  f"mp4mux ! filesink name=sink location=output.mp4"
160
  )
161
  pipeline = Gst.parse_launch(pipeline_str)
@@ -231,16 +235,28 @@ bus.timed_pop_filtered(
231
  pipeline.set_state(Gst.State.NULL)
232
  ```
233
 
234
- Expected output with the sample video and the zone/threshold above:
 
 
235
 
236
  ```text
237
- LOITERING id=2 dwell=1.6s anchor=(529,258)
238
- LOITERING id=9 dwell=1.6s anchor=(527,250)
 
 
239
  ```
240
 
 
 
241
  The annotated video is saved to `output.mp4` with green bounding boxes and
242
  track IDs drawn by `gvawatermark` around every detected person.
243
 
 
 
 
 
 
 
244
  ---
245
 
246
  ## License
 
6
  |---|---|
7
  | **Category** | Object Detection + Tracking + Zone Analytics |
8
  | **Source Framework** | PyTorch (Ultralytics) |
9
+ | **Supported Precisions** | FP32, FP16, INT8 (mixed-precision) |
10
  | **Inference Engine** | OpenVINO |
11
  | **Hardware** | CPU, GPU, NPU |
12
  | **Detected Class** | `person` (COCO class 0) |
 
44
  source .venv/bin/activate
45
  ```
46
 
47
+ > **Note:** The `--system-site-packages` flag is required so the virtual
48
+ > environment can access the system-installed OpenVINO and DLStreamer Python
49
+ > packages.
50
+
51
  ---
52
 
53
  ## Getting Started
 
159
  f"gvametaconvert add-empty-results=true ! queue ! "
160
  f"gvafpscounter ! "
161
  f"gvawatermark ! videoconvert ! video/x-raw,format=I420 ! "
162
+ f"openh264enc ! h264parse ! "
163
  f"mp4mux ! filesink name=sink location=output.mp4"
164
  )
165
  pipeline = Gst.parse_launch(pipeline_str)
 
235
  pipeline.set_state(Gst.State.NULL)
236
  ```
237
 
238
+ Expected output with the sample video and the zone/threshold above
239
+ (exact track IDs and anchor coordinates may vary between runs due to
240
+ tracker non-determinism):
241
 
242
  ```text
243
+ LOITERING id=26 dwell=5.0s anchor=(147,341)
244
+ LOITERING id=27 dwell=5.0s anchor=(122,337)
245
+ LOITERING id=29 dwell=5.0s anchor=(90,322)
246
+ ...
247
  ```
248
 
249
+ Approximately 10–12 loitering events are expected over the full video.
250
+
251
  The annotated video is saved to `output.mp4` with green bounding boxes and
252
  track IDs drawn by `gvawatermark` around every detected person.
253
 
254
+ > **Known warning:** The `openh264enc` element prints
255
+ > `[OpenH264] this = 0x..., Error:CWelsH264SVCEncoder::EncodeFrame(), cmInitParaError.`
256
+ > on the first frame. This is a benign initialization message — the output
257
+ > video is encoded correctly. The warning comes from the OpenH264 library's
258
+ > internal logging and does not indicate a real error.
259
+
260
  ---
261
 
262
  ## License