1.2 导出检测结果 | 1. SuperVision介绍 | ImGui-WebAssembly

## 导出检测结果

```text 
「supervision book」是一本开源电子书，PDF/随书代码/资源下载： https://github.com/ThisisGame/supervision_book 
```

可以导出物体检测结果到csv、json、自定义结构。

官方文档 `https://supervision.roboflow.com/latest/how_to/save_detections/`

### 准备视频

先下载测试视频。

```python
#file:files\supervision\download_people_walking_mp4.py

from supervision.assets import download_assets, VideoAssets

download_assets(VideoAssets.PEOPLE_WALKING)
```

### 导出到csv

```python
#file:files\supervision\save_detections_to_csv.py

import supervision as sv
from ultralytics import YOLO

model = YOLO("yolov8n.pt")
frames_generator = sv.get_video_frames_generator("people-walking.mp4")

with sv.CSVSink("./save_detactions_to_csv.csv") as sink:
    for frame in frames_generator:

results = model(frame)[0]
        detections = sv.Detections.from_ultralytics(results)
        sink.append(detections, {})
```

执行后就会将检测结果输出到`./save_detactions_to_csv.csv`，如下图：

![](md/supervision_book/imgs/supervision/save_detections_to_csv.png)<a id="antiCollectorAdTxt" href="https://github.com/ThisisGame/supervision_book">「supervision book」是一本开源电子书，PDF/随书代码/资源下载： https://github.com/ThisisGame/supervision_book</a>

### 导出自定义列

上面直接导出的csv，缺少了帧序号，下面就用自定义列来添加帧序号。

```python
#file:files\supervision\save_detections_to_csv_custom_column.py

import supervision as sv
from ultralytics import YOLO

model = YOLO("yolov8n.pt")
frames_generator = sv.get_video_frames_generator("people-walking.mp4")

with sv.CSVSink("./save_detections_to_csv_custom_column.csv") as sink:
    for frame_index, frame in enumerate(frames_generator):

results = model(frame)[0]
        detections = sv.Detections.from_ultralytics(results)
        sink.append(detections, {"frame_index": frame_index})#用自定义列来添加帧序号
```

![](md/supervision_book/imgs/supervision/save_detections_to_csv_custom_column.png)

### 导出到Json

```python
#file:files\supervision\save_detections_to_json.py

import supervision as sv
from ultralytics import YOLO

model = YOLO("yolov8n.pt")
frames_generator = sv.get_video_frames_generator("people-walking.mp4")

with sv.JSONSink("./save_detections_to_json.json") as sink:
    for frame_index, frame in enumerate(frames_generator):

results = model(frame)[0]
        detections = sv.Detections.from_ultralytics(results)
        sink.append(detections, {"frame_index": frame_index})
```

![](md/supervision_book/imgs/supervision/save_detections_to_json.png)