mirror of
https://github.com/orbbec/OrbbecSDK_ROS2.git
synced 2026-10-08 22:07:46 +08:00
docs: document side-stream benchmark tools
This commit is contained in:
ob-yalian
@@ -4,17 +4,18 @@ This section introduces the performance benchmark tools, their purpose, and the
|
||||
|
||||
## common_benchmark_node.py
|
||||
|
||||
`common_benchmark_node.py` monitors Orbbec camera performance in a ROS environment. It collects and records key camera metrics such as frame rate, latency, system resource usage, and packet/drop rate to help evaluate camera node stability and performance. Statistics are updated once per second.
|
||||
`common_benchmark_node.py` monitors Orbbec camera performance in a ROS environment. It collects and records subscriber-side metrics such as image frame rate, latency, system resource usage, and estimated frame loss to help evaluate camera node stability and performance. Statistics are updated once per second.
|
||||
|
||||
Features:
|
||||
|
||||
- Measures published image frame rate and latency (current, minimum, maximum, and average).
|
||||
- Measures the frame rate and latency of images received by the subscriber (current, minimum, maximum, and average).
|
||||
- Monitors camera node CPU/ARM usage (current, minimum, maximum, and average).
|
||||
- Tracks frame drop rate (publisher) and packet loss rate (subscriber).
|
||||
- Estimates subscriber-side frame loss from image message timestamps.
|
||||
- Supports `color`, `left_color`, `right_color`, `depth`, `ir`, `left_ir`, and `right_ir` image streams, as well as compressed image and point cloud topics.
|
||||
- Prints real-time statistics at 1 Hz and saves results to a CSV file.
|
||||
- Supports configurable runtime and CSV output path.
|
||||
|
||||
In ROS 1, both frame drop rate and packet loss rate can be measured. In ROS 2, the header does not contain the `seq` field, so only publisher-side frame drop rate is calculated.
|
||||
In ROS 2, the node estimates subscriber-side frame loss from adjacent image message timestamps. When `--ideal_fps` is set, it uses the specified ideal frame rate; otherwise, it learns the nominal frame interval from received image timestamps.
|
||||
|
||||
|
||||
|
||||
@@ -32,6 +33,9 @@ Parameters:
|
||||
|
||||
- `--run_time`: Monitoring duration, specified as a time string such as `"10s"`, `"5m"`, `"1h"`, or `"2d"`. The default is 10 seconds.
|
||||
- `--csv_file`: Output CSV file path. By default, it is saved in the workspace directory as `camera_monitor_log.csv`.
|
||||
- `--ideal_fps`: Ideal frame rate used for subscriber-side frame-loss estimation. A value greater than `0` overrides the interval learned from image timestamps.
|
||||
- `--camera_names`: Camera namespaces to monitor, separated by commas.
|
||||
- `--topics`: Topics to monitor, separated by commas. Use raw stream names such as `color,left_color,right_color`, or full raw/compressed image and point cloud topics. Automatic discovery selects only raw image topics; compressed images and point clouds require full topic names.
|
||||
|
||||
Multi-camera monitoring example:
|
||||
|
||||
@@ -39,7 +43,8 @@ Multi-camera monitoring example:
|
||||
ros2 run orbbec_camera common_benchmark_node.py \
|
||||
--run_time 1h \
|
||||
--csv_file /tmp/cam_log.csv \
|
||||
--camera_names camera_01,camera_02
|
||||
--camera_names camera_01,camera_02 \
|
||||
--topics color,left_color,right_color
|
||||
```
|
||||
|
||||
## service_benchmark_node.py
|
||||
@@ -205,7 +210,7 @@ Output data files are stored in the `ob_benchmark` folder with names such as `0.
|
||||
|
||||
## start_benchmark_node
|
||||
|
||||
`start_benchmark_node` is the subscriber used in the benchmark workflow. It subscribes to multi-camera color, depth, IR, and point cloud topics according to `camera_name` in `start_benchmark_params.json`. It is usually used together with benchmark launch files.
|
||||
`start_benchmark_node` is the subscriber used in the benchmark workflow. It subscribes to multi-camera `color`, `left_color`, `right_color`, `depth`, `left_ir`, `right_ir`, and point cloud topics according to `camera_name` in `start_benchmark_params.json`. It is usually used together with benchmark launch files.
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera start_benchmark_node
|
||||
|
||||
@@ -4,7 +4,7 @@ This section describes multi-camera image saving, synchronization verification,
|
||||
|
||||
## multi_save_rgbir_node
|
||||
|
||||
`multi_save_rgbir_node` subscribes to multi-camera RGB/IR images and metadata according to `multi_save_rgbir_params.json`, and uses the `start_capture` service to trigger saving. Start the corresponding multi-camera nodes before using it.
|
||||
`multi_save_rgbir_node` subscribes to multi-camera RGB/IR images and metadata according to `multi_save_rgbir_params.json`, and uses the `start_capture` service to trigger saving. Supported stream names are `color`, `left_color`, `right_color`, `ir`, `left_ir`, and `right_ir`. Start the corresponding multi-camera nodes before using it.
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera multi_save_rgbir_node
|
||||
@@ -16,6 +16,14 @@ The configuration file is:
|
||||
orbbec_camera/config/tools/multisavergbir/multi_save_rgbir_params.json
|
||||
```
|
||||
|
||||
Set the streams to save in `stream_names`, for example:
|
||||
|
||||
```json
|
||||
"stream_names": ["color", "left_color", "right_color"]
|
||||
```
|
||||
|
||||
When `stream_names` is an empty array, the node automatically discovers stable available streams after startup. When streams are configured explicitly, every selected stream must receive the requested number of frames before capture completes.
|
||||
|
||||
Trigger saving 10 frames:
|
||||
|
||||
```bash
|
||||
@@ -26,7 +34,7 @@ ros2 service call /start_capture orbbec_camera_msgs/srv/SetInt32 "{data: 10}"
|
||||
|
||||
`image_sync_example_node` verifies multi-image timestamp synchronization online. It subscribes to 1 to 8 image topics, displays synchronized images, and prints timestamp difference and FPS statistics. Start the camera node before using it.
|
||||
|
||||
If `sync_topics` is not set, the node automatically discovers color/depth image topics:
|
||||
If `sync_topics` is not set, the node automatically discovers image topics for `left_color`, `right_color`, `left_ir`, `right_ir`, `color`, `depth`, and `ir`. The node supports synchronizing up to 8 image topics:
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera image_sync_example_node
|
||||
|
||||
@@ -4,17 +4,18 @@
|
||||
|
||||
## common_benchmark_node.py
|
||||
|
||||
`common_benchmark_node.py` 是一个用于监控在 ROS 环境中运行的 Orbbec 相机性能的工具。它实时收集和记录关键相机指标，如帧率、延迟、系统资源使用和丢包率，帮助用户评估相机节点的稳定性和性能（每秒更新一次）。
|
||||
`common_benchmark_node.py` 是一个用于监控在 ROS 环境中运行的 Orbbec 相机性能的工具。它在订阅端实时收集和记录关键相机指标，如图像帧率、延迟、系统资源使用和估算丢帧率，帮助用户评估相机节点的稳定性和性能（每秒更新一次）。
|
||||
|
||||
功能：
|
||||
|
||||
- 测量发布的图像帧率和延迟（当前、最小、最大、平均）。
|
||||
- 测量订阅端接收的图像帧率和延迟（当前、最小、最大、平均）。
|
||||
- 监控相机节点的 CPU/ARM 使用率（当前、最小、最大、平均）。
|
||||
- 跟踪丢帧率（发布者）和丢包率（订阅者）。
|
||||
- 根据图像消息时间戳估算订阅端丢帧率。
|
||||
- 支持 `color`、`left_color`、`right_color`、`depth`、`ir`、`left_ir` 和 `right_ir` 图像流，也支持压缩图像和点云 topic。
|
||||
- 将实时统计信息（1 Hz）打印到终端并将结果保存到 CSV 文件。
|
||||
- 支持可配置的运行时长和 CSV 输出路径。
|
||||
|
||||
在 ROS1 中，可以测量丢帧率和丢包率，而在 ROS2 中，header 缺少 `seq` 字段，因此仅计算发布者端的丢帧率。
|
||||
在 ROS2 中，节点根据相邻图像消息的时间戳估算订阅端丢帧率。设置 `--ideal_fps` 时使用指定的理想帧率；未设置时根据收到的图像时间戳学习名义帧间隔。
|
||||
|
||||
|
||||
|
||||
@@ -32,6 +33,9 @@ ros2 run orbbec_camera common_benchmark_node.py \
|
||||
|
||||
- `--run_time`：监控持续时间，指定为时间字符串，如 `"10s"`、`"5m"`、`"1h"`、`"2d"`。默认为 10 秒。
|
||||
- `--csv_file`：输出 CSV 文件的路径。默认情况下，它保存在工作空间目录中，名称为 `camera_monitor_log.csv`。
|
||||
- `--ideal_fps`：订阅端丢帧估算使用的理想帧率。大于 `0` 时覆盖根据图像时间戳学习的帧间隔。
|
||||
- `--camera_names`：要监控的相机命名空间，多个名称用逗号分隔。
|
||||
- `--topics`：要监控的 topic，多个值用逗号分隔。可以填写原始流名称（例如 `color,left_color,right_color`），也可以填写完整的原始/压缩图像或点云 topic。自动发现只选择原始图像 topic；压缩图像和点云需要填写完整 topic 名称。
|
||||
|
||||
多相机监控示例：
|
||||
|
||||
@@ -39,7 +43,8 @@ ros2 run orbbec_camera common_benchmark_node.py \
|
||||
ros2 run orbbec_camera common_benchmark_node.py \
|
||||
--run_time 1h \
|
||||
--csv_file /tmp/cam_log.csv \
|
||||
--camera_names camera_01,camera_02
|
||||
--camera_names camera_01,camera_02 \
|
||||
--topics color,left_color,right_color
|
||||
```
|
||||
|
||||
## service_benchmark_node.py
|
||||
@@ -207,7 +212,7 @@ ros2 run orbbec_camera ob_benchmark_node
|
||||
|
||||
## start_benchmark_node
|
||||
|
||||
`start_benchmark_node` 是 benchmark 流程中的订阅端，会按 `start_benchmark_params.json` 中的 `camera_name` 订阅多相机 color、depth、IR 和 point cloud topic。它通常与 benchmark launch 配合使用。
|
||||
`start_benchmark_node` 是 benchmark 流程中的订阅端，会按 `start_benchmark_params.json` 中的 `camera_name` 订阅多相机 `color`、`left_color`、`right_color`、`depth`、`left_ir`、`right_ir` 和 point cloud topic。它通常与 benchmark launch 配合使用。
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera start_benchmark_node
|
||||
|
||||
@@ -4,7 +4,7 @@
|
||||
|
||||
## multi_save_rgbir_node
|
||||
|
||||
`multi_save_rgbir_node` 根据 `multi_save_rgbir_params.json` 配置订阅多相机 RGB/IR 图像和 metadata，并通过 `start_capture` 服务触发保存。使用前需要先启动对应的多相机节点。
|
||||
`multi_save_rgbir_node` 根据 `multi_save_rgbir_params.json` 配置订阅多相机 RGB/IR 图像和 metadata，并通过 `start_capture` 服务触发保存。支持的流名称为 `color`、`left_color`、`right_color`、`ir`、`left_ir` 和 `right_ir`。使用前需要先启动对应的多相机节点。
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera multi_save_rgbir_node
|
||||
@@ -16,6 +16,14 @@ ros2 run orbbec_camera multi_save_rgbir_node
|
||||
orbbec_camera/config/tools/multisavergbir/multi_save_rgbir_params.json
|
||||
```
|
||||
|
||||
在配置文件的 `stream_names` 中填写要保存的流名称，例如：
|
||||
|
||||
```json
|
||||
"stream_names": ["color", "left_color", "right_color"]
|
||||
```
|
||||
|
||||
`stream_names` 为空数组时，节点会在启动后自动发现稳定可用的流。配置流名称后，触发保存时需要每个指定流都收到所需帧数。
|
||||
|
||||
触发保存 10 帧：
|
||||
|
||||
```bash
|
||||
@@ -26,7 +34,7 @@ ros2 service call /start_capture orbbec_camera_msgs/srv/SetInt32 "{data: 10}"
|
||||
|
||||
`image_sync_example_node` 用于在线验证多路图像时间戳同步情况。它会订阅 1 到 8 路图像 topic，显示同步图像，并输出时间戳差和 FPS 统计。使用前需要先启动相机节点。
|
||||
|
||||
未设置 `sync_topics` 时，节点会自动发现 color/depth 图像 topic：
|
||||
未设置 `sync_topics` 时，节点会自动发现以下图像流的 topic：`left_color`、`right_color`、`left_ir`、`right_ir`、`color`、`depth` 和 `ir`。节点最多同步 8 路图像：
|
||||
|
||||
```bash
|
||||
ros2 run orbbec_camera image_sync_example_node
|
||||
|
||||
Reference in New Issue
Block a user