|
@@ -13,7 +13,7 @@ This results in a MJPEG stream with objects identified that has a lower latency
|
|
|
## Getting Started
|
|
|
Build the container with
|
|
|
```
|
|
|
-docker build -t realtime-od .
|
|
|
+docker build -t frigate .
|
|
|
```
|
|
|
|
|
|
Download a model from the [zoo](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md).
|
|
@@ -31,7 +31,7 @@ docker run --rm \
|
|
|
-e MQTT_HOST='your.mqtthost.com' \
|
|
|
-e MQTT_TOPIC_PREFIX='cameras/1' \
|
|
|
-e DEBUG='0' \
|
|
|
-realtime-od:latest
|
|
|
+frigate:latest
|
|
|
```
|
|
|
|
|
|
Example compose:
|
|
@@ -39,7 +39,7 @@ Example compose:
|
|
|
frigate:
|
|
|
container_name: frigate
|
|
|
restart: unless-stopped
|
|
|
- image: realtime-od:latest
|
|
|
+ image: frigate:latest
|
|
|
volumes:
|
|
|
- <path_to_frozen_detection_graph.pb>:/frozen_inference_graph.pb:ro
|
|
|
- <path_to_labelmap.pbtext>:/label_map.pbtext:ro
|