# 使用 Neural Processing SDK 进行目标检测和编码

Source: [https://docs.qualcomm.com/doc/80-70015-50SC/topic/single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd.html](https://docs.qualcomm.com/doc/80-70015-50SC/topic/single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd.html)

这些用例使用 yolonas.dlc 目标检测模型和 Qualcomm Neural Processing SDK 来识别摄像头流中的目标对象，在检测到的目标对象上叠加或合成边框，然后将数据流编码为 H.264 码流。

## 变体 1：使用 qtioverlay 插件应用检测叠加

运行用例：

    setprop persist.overlay.use_c2d_blit 2Copy to clipboard

    gst-launch-1.0 -e \
    qtiqmmfsrc name=camsrc ! video/x-raw\(memory:GBM\),format=NV12,width=1280,height=720,framerate=30/1,compression=ubwc ! queue ! tee name=split \
    split. ! queue ! qtimetamux name=metamux ! queue ! qtioverlay ! queue ! v4l2h264enc capture-io-mode=5 output-io-mode=5 ! h264parse ! queue ! mp4mux ! queue ! filesink location=/opt/video.mp4 \
    split. ! queue ! qtimlvconverter ! queue ! qtimlsnpe delegate=dsp model=/opt/yolonas.dlc layers="</heads/Mul, /heads/Sigmoid>" ! queue ! qtimlvdetection threshold=51.0 results=10 module=yolo-nas labels=/opt/yolonas.labels ! text/x-raw ! queue ! metamux.Copy to clipboard

如需停止用例，可按下 CTRL + C。

Figure : 边框叠加和编码的 pipeline
                ![](data:image/png;base64,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)

下图显示了用例执行流程：

1. 从摄像头源传来的视频流中识别目标对象场景。
2. 使用 overlaylib 将边框叠加在检测到的目标对象上。
3. 将此数据流编码为 H.264 码流。
4. 将数据流多路复用到 MP4 容器中并存储为 MP4 文件。

pipeline 执行的顺序处理阶段如下表所示：

| 过程 | 说明 |
| --- | --- |
| [qtiqmmfsrc](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtiqmmfsrc.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_l2f_zgm_vbc"><br>                                    <li class="li">采集视频流（源）并创建源的两个副本：<ul class="ul" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_m2f_zgm_vbc"><br>                                            <li class="li">一个视频流被发送到 qtimetamux 插件以保留该视频流。</li><br><br>                                            <li class="li">另一个视频流被发送到 ML 推理 pipeline。</li><br><br>                                        </ul><br></li><br><br>                                </ol> |
| **预处理** | **预处理** |
| [qtimlvconverter](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlvconverter.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_xsf_q5l_vbc"><br>                                    <li class="li">在其接收设备接插口上接收视频流。</li><br><br>                                    <li class="li">执行预处理：<ul class="ul" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ul_ff2_twl_vbc"><br>                                            <li class="li">色彩转换</li><br><br>                                            <li class="li">缩小/放大</li><br><br>                                            <li class="li">在模型使用浮点值输入时，对流数据进行归一化</li><br><br>                                        </ul><br></li><br><br>                                    <li class="li">在其源接插口上将视频流转换为张量数据。<p class="p">目标检测模型使用此张量数据进行推理。</p><br></li><br><br>                                </ol> |
| **推理** | **推理** |
| [qtimlsnpe](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlsnpe.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_ufn_2lm_vbc"><br>                                    <li class="li">加载目标检测模型。</li><br><br>                                    <li class="li">为选择的 delegate 修改图形。</li><br><br>                                    <li class="li">在其接收设备接插口上接收张量数据。</li><br><br>                                    <li class="li">运行推理并在其源接插口上生成带有目标检测结果的张量数据。</li><br><br>                                </ol> |
| **后处理** | **后处理** |
| [qtimlvdetection](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlvdetection.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_ky5_grn_vbc"><br>                                    <li class="li">接收来自目标检测模型的推理张量。</li><br><br>                                    <li class="li">将接收设备接插口上的推理张量转换为视频或文本等格式，稍后可由多媒体插件进行处理。</li><br><br>                                    <li class="li">将阈值应用于所选的结果数。</li><br><br>                                    <li class="li">加载检测模型的相应模块。 <p class="p">在此用例中，qtimlvdetection 执行以下操作：</p><ol class="ol" type="a" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_jcd_wnk_5bc"><br>                                            <li class="li">加载 YOLO-NAS 子模块。</li><br><br>                                            <li class="li">将结果生成为文本结构。</li><br><br>                                            <li class="li">接着发送到 qtimetamux 的接收设备接插口。</li><br><br>                                        </ol><br></li><br><br>                                </ol> |
| [qtimetamux](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimetamux.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_ll3_x5l_vbc"><br>                                    <li class="li">在接收设备接插口上接收视频流和文本流，以及与视频流相对应的边框结果。</li><br><br>                                    <li class="li">使用接收设备接插口中的视频流内容生成 GST 缓冲区。</li><br><br>                                    <li class="li">将边框作为 <code class="ph codeph">GstVideoRegionOfInterest</code> 从数据接收设备接插口添加到其源接插口上的 GST 缓冲区元数据（元多路复用）。</li><br><br>                                </ol> |
| [qtioverlay](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtioverlay.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_wst_y5l_vbc"><br>                                    <li class="li">接收多路复用流。</li><br><br>                                    <li class="li">使用 CL 将边框叠加在 VideoFrame 上。</li><br><br>                                    <li class="li">在其源接插口上生成带有叠加层的 GST 缓冲区。</li><br><br>                                </ol> |
| [v4l2h264enc](https://docs.qualcomm.com/doc/80-70015-50SC/topic/v4l2h264enc.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_wsc_bsn_vbc"><br>                                    <li class="li">将参数应用于在接收设备上接收的视频流的每一帧。</li><br><br>                                    <li class="li">将其编码为码流，并通过其源接插口发送。</li><br><br>                                </ol> |
| h264parse | 将与码流对应的其他信息添加到 GStreamer 缓冲区元数据。 |
| mp4mux | 接收这些缓冲区并创建具有格式规范缓冲区的容器。 |
| **输出** | **输出** |
| Filesink | 将生成的数据流存储在 /opt/video.mp4 文件中。 |
| Playback | 从主机拉取 video.mp4 并在媒体播放器上播放：<br>`scp root@<IP address of target device>:/opt/ <destination directory>` |

## 变体 2：使用 qtivcomposer 混合原始帧与检测掩码

运行用例：

    gst-launch-1.0 -e \
    qtiqmmfsrc name=camsrc ! video/x-raw\(memory:GBM\),format=NV12,width=1280,height=720,framerate=30/1,compression=ubwc ! queue ! tee name=split \
    split. ! queue ! qtivcomposer name=mixer ! queue ! video/x-raw\(memory:GBM\),format=NV12,width=1920,height=1080,interlace-mode=progressive,colorimetry=bt601 ! v4l2h264enc capture-io-mode=5 output-io-mode=5 ! h264parse ! queue ! mp4mux ! queue ! filesink location=/opt/video.mp4 \
    split. ! queue ! qtimlvconverter ! queue ! qtimlsnpe delegate=dsp model=/opt/yolonas.dlc layers="</heads/Mul, /heads/Sigmoid>" ! queue ! qtimlvdetection threshold=51.0 results=10 module=yolo-nas labels=/opt/yolonas.labels ! video/x-raw,width=640,height=360 ! queue ! mixer.Copy to clipboard

如需停止用例，可按下 CTRL + C。

Figure : 使用 qtivcomposer 进行边框掩码和编码的 pipeline
                ![](data:image/png;base64,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)

下图显示了用例执行流程：

1. 从摄像头源传来的视频流中识别目标对象场景。
2. 使用 qtivcomposer 在检测到的目标对象和原始视频流上合成边框。
3. 将此数据流编码为 H.264 码流。
4. 将数据流多路复用到 MP4 容器中并存储为 MP4 文件。

pipeline 执行的顺序处理阶段如下表所示：

| 过程 | 说明 |
| --- | --- |
| [qtiqmmfsrc](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtiqmmfsrc.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_wqx_ntn_vbc"><br>                                    <li class="li">采集视频流（源）并创建源的两个副本：<ul class="ul" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_xqx_ntn_vbc"><br>                                            <li class="li">一个视频流被发送到 qtimetamux 插件以保留该视频流。</li><br><br>                                            <li class="li">另一个视频流被发送到 ML 推理 pipeline。</li><br><br>                                        </ul><br></li><br><br>                                </ol> |
| **预处理** | **预处理** |
| [qtimlvconverter](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlvconverter.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_yqx_ntn_vbc"><br>                                    <li class="li">在其接收设备接插口上接收视频流。</li><br><br>                                    <li class="li">执行预处理：<ul class="ul" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ul_zqx_ntn_vbc"><br>                                            <li class="li">色彩转换</li><br><br>                                            <li class="li">缩小/放大</li><br><br>                                            <li class="li">在模型使用浮点值输入时，对流数据进行归一化</li><br><br>                                        </ul><br></li><br><br>                                    <li class="li">在其源接插口上将视频流转换为张量数据。<p class="p">目标检测模型使用此张量数据进行推理。</p><br></li><br><br>                                </ol> |
| **推理** | **推理** |
| [qtimlsnpe](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlsnpe.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_arx_ntn_vbc"><br>                                    <li class="li">加载目标检测模型。</li><br><br>                                    <li class="li">为选择的 delegate 修改图形。</li><br><br>                                    <li class="li">在其接收设备接插口上接收张量数据。</li><br><br>                                    <li class="li">运行推理并在其源接插口上生成带有目标检测结果的张量数据。</li><br><br>                                </ol> |
| **后处理** | **后处理** |
| [qtimlvdetection](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtimlvdetection.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_brx_ntn_vbc"><br>                                    <li class="li">接收来自目标检测模型的推理张量。</li><br><br>                                    <li class="li">将接收设备接插口上的推理张量转换为视频或文本等格式，稍后可由多媒体插件进行处理。</li><br><br>                                    <li class="li">将阈值应用于所选的结果数。</li><br><br>                                    <li class="li">加载检测模型的相应模块。 <p class="p">在此用例中，qtimlvdetection 执行以下操作：</p><ol class="ol" type="a" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_crx_ntn_vbc"><br>                                            <li class="li">加载 YOLO-NAS 子模块。</li><br><br>                                            <li class="li">生成仅包含可叠加在目标对象上的边框的视频帧。</li><br><br>                                            <li class="li">将其发送到 qtivcomposer 的接收设备接插口。</li><br><br>                                        </ol><br></li><br><br>                                </ol> |
| [qtivcomposer](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qtivcomposer.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_wgh_rtn_vbc"><br>                                    <li class="li">在其接收设备接插口上接收原始视频流和带有边框的视频流</li><br><br>                                    <li class="li">在其源接插口上，生成由从接收设备接插口处理的视频流组成的内容。</li><br><br>                                </ol> |
| [v4l2h264enc](https://docs.qualcomm.com/doc/80-70015-50SC/topic/v4l2h264enc.html) | <ol class="ol" id="single-camera-stream-with-object-detection-and-encode-with-mobilenet-v2-ssd__ol_frx_ntn_vbc"><br>                                    <li class="li">将参数应用于在接收设备接插口上接收的视频流的每一帧。</li><br><br>                                    <li class="li">将其编码为码流，并通过其源接插口发送。</li><br><br>                                </ol> |
| h264parse | 将与码流对应的其他信息添加到 GStreamer 缓冲区元数据。 |
| mp4mux | 接收这些缓冲区并创建具有格式规范缓冲区的容器。 |
| **输出** | **输出** |
| Filesink | 将生成的数据流存储在 /opt/video.mp4 文件中。 |
| Playback | 从主机拉取 video.mp4 并在媒体播放器上播放：<br>`scp root@<IP address of target device>:/opt/ <destination directory>` |

**Parent Topic:** [Qualcomm Neural Processing SDK 用例](https://docs.qualcomm.com/doc/80-70015-50SC/topic/qualcomm-neural-processing-sdk-use-cases.html)

Last Published: Nov 11, 2025

[Previous Topic
使用 Neural Processing SDK 进行目标检测和显示](https://docs.qualcomm.com/bundle/publicresource/80-70015-50SC/topics/single-camera-stream-with-object-detection-and-display-with-mobilenet-v2-ssd.md) [Next Topic
使用 Neural Processing SDK 进行图像分割和显示](https://docs.qualcomm.com/bundle/publicresource/80-70015-50SC/topics/single-camera-stream-with-image-segmentation-and-display-with-deeplabv3-quantized.md)