# 使用 Neural Processing SDK 进行图像分类和显示 

Source: [https://docs.qualcomm.com/doc/80-70017-50SC/topic/single-camera-stream-with-image-classification-and-display-with-mobilenet-v1.html](https://docs.qualcomm.com/doc/80-70017-50SC/topic/single-camera-stream-with-image-classification-and-display-with-mobilenet-v1.html)

这些用例使用带有 Qualcomm Neural Processing SDK 的 Inceptionv3 模型对场景进行分类，叠加或合成分类标签，接着显示结果。

可以使用 TensorFlow 获取任何公开可用的分类模型，并将其转换为 `.dlc` 格式，如 [TensorFlow 模型转换](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-2/model_conv_tensorflow.html)中所述。

## 使用 qtioverlay 插件来应用分类叠加

运行用例：

    gst-launch-1.0 -e --gst-debug=2 \
    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 ! waylandsink fullscreen=true sync=false \
    split. ! queue ! qtimlvconverter ! queue ! qtimlsnpe delegate=dsp model=/opt/inceptionv3.dlc ! queue ! \
    qtimlvclassification threshold=40.0 results=2 module=mobilenet labels=/opt/classification.labels ! text/x-raw ! queue ! metamux.Copy to clipboard

如需停止用例，可按下 CTRL + C。

Figure : 分类叠加 pipeline
                
                ![](data:image/png;base64,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)

下图显示了用例执行流程：

- 对从摄像头源传来的视频流中的场景进行分类。
- 使用 overlaylib 叠加分类标签。
- 在本地显示屏上显示结果。

pipeline 执行的顺序处理阶段如下表所示：

| 处理过程 | 说明 |
| --- | --- |
| [qtiqmmfsrc](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtiqmmfsrc.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__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-image-classification-and-display-with-mobilenet-v1__ol_m2f_zgm_vbc"><br>                                            <li class="li">一个视频流被发送到 qtimetamux 插件以保留视频流。</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/80-70017-50SC/topic/qtimlvconverter.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_xsf_q5l_vbc"><br>                                    <li class="li">在其接收端上接收视频流。</li><br><br>                                    <li class="li">执行预处理：<ul class="ul" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ul_ff2_twl_vbc"><br>                                            <li class="li">颜色转换</li><br><br>                                            <li class="li">缩小/放大</li><br><br>                                            <li class="li">在模型需要浮点值作为输入时，对流数据进行归一化</li><br><br>                                        </ul><br></li><br><br>                                    <li class="li">在其发送端上将视频流转换为张量数据。<p class="p">分类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/80-70017-50SC/topic/qtimlsnpe.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_bwn_s5l_vbc"><br>                                    <li class="li">加载模型。</li><br><br>                                    <li class="li">为选择的 delegate 修改图。</li><br><br>                                    <li class="li">在其接收端上接收张量数据。</li><br><br>                                    <li class="li">执行推理并在其发送端上生成包含推理结果的张量数据。</li><br><br>                                </ol> |
| **后处理** | **后处理** |
| [qtimlvclassification](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtimlvclassification.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_gr1_w5l_vbc"><br>                                    <li class="li">从其接收端上的模型接收推理张量。</li><br><br>                                    <li class="li">将张量转换为多媒体插件稍后可以处理的视频或文本等格式。</li><br><br>                                    <li class="li">将阈值应用于所选的结果数。</li><br><br>                                    <li class="li">加载分类模型的相应模块。 <p class="p">在此用例中，qtimlvclassification 执行以下操作：</p><ol class="ol" type="a" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_rrb_1xl_vbc"><br>                                            <li class="li">加载模型的子模块。</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 的接收端。</li><br><br>                                        </ol><br></li><br><br>                                </ol> |
| [qtimetamux](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtimetamux.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_ll3_x5l_vbc"><br>                                    <li class="li">接收视频流和文本流，其接收端上的分类结果与视频流相对应。</li><br><br>                                    <li class="li">在其接收端上生成包含视频流内容的 GST 缓冲存。</li><br><br>                                    <li class="li">将数据接收端上的分类结果添加到其发送端上的 GST 缓存元数据（元数据多路复用）。</li><br><br>                                </ol> |
| [qtioverlay](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtioverlay.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_wst_y5l_vbc"><br>                                    <li class="li">接收多路复用流。</li><br><br>                                    <li class="li">使用 CL 将分类标签叠加在 VideoFrame 上。 </li><br><br>                                    <li class="li">在其发送端上生成带有叠加层的 GST 缓存。</li><br><br>                                </ol> |
| **输出** | **输出** |
| [Waylandsink](https://docs.qualcomm.com/doc/80-70017-50SC/topic/waylandsink.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_fd3_wc5_vbc"><br>                                    <li class="li">在其接收端上接收视频流。</li><br><br>                                    <li class="li">将视频流提交到 Weston。 </li><br><br>                                    <li class="li">Weston 在本地显示设备上渲染视频流和为该场景生成的可能分类。</li><br><br>                                </ol> |

## 使用 qtivcomposer 混合原始帧与分类掩码

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 ! qtimetamux name=metamux ! queue ! qtioverlay ! queue ! waylandsink fullscreen=true sync=false \
    split. ! queue ! qtimlvconverter ! queue ! qtimlsnpe delegate=dsp model=/opt/inceptionv3.dlc ! queue ! qtimlvclassification threshold=40.0 results=2 module=mobilenet labels=/opt/classification.labels ! text/x-raw ! queue ! metamux.Copy to clipboard

如需停止用例，可按下 CTRL + C。

Figure : 使用 qtivcomposer 进行分类的 pipeline
                
                ![](data:image/png;base64,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)

下图显示了用例执行流程：
- 对从摄像头源传来的视频流中的场景进行分类。
- 使用 qtivcomposer 合成分类标签和视频流。
- 在本地显示屏上显示结果。

pipeline 执行的顺序处理阶段如下表所示：

| 处理过程 | 说明 |
| --- | --- |
| [qtiqmmfsrc](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtiqmmfsrc.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_x5l_jd5_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-image-classification-and-display-with-mobilenet-v1__ul_n44_nwl_vbc"><br>                                            <li class="li">一个视频流被发送到 qtivcomposer 插件以保留视频流。</li><br><br>                                            <li class="li">另一个流被发送至 pipeline 中的 ML 推理分支。</li><br><br>                                        </ul><br></li><br><br>                                </ol> |
| **预处理** | **预处理** |
| [qtimlvconverter](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtimlvconverter.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_i5w_4wl_vbc"><br>                                    <li class="li">在其接收端上接收视频流。</li><br><br>                                    <li class="li">执行预处理：<ul class="ul" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_zdw_qwl_vbc"><br>                                            <li class="li">颜色转换</li><br><br>                                            <li class="li">缩小/放大</li><br><br>                                            <li class="li">在模型需要浮点值作为输入时，对流数据进行归一化</li><br><br>                                        </ul><br></li><br><br>                                    <li class="li">在其发送端上将视频流转换为张量数据。<p class="p">分类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/80-70017-50SC/topic/qtimlsnpe.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_u1l_cxl_vbc"><br>                                    <li class="li">加载模型。</li><br><br>                                    <li class="li">为选择的 delegate 修改图。</li><br><br>                                    <li class="li">在其接收端上接收张量数据。</li><br><br>                                    <li class="li">执行推理并在其发送端上生成包含推理结果的张量数据。</li><br><br>                                </ol> |
| **后处理** | **后处理** |
| [qtimlvclassification](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtimlvclassification.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_o3v_2xl_vbc"><br>                                    <li class="li">从其接收端上的模型接收推理结果。 </li><br><br>                                    <li class="li">将推理张量转换为视频或文本等格式，稍后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">加载分类模型的相应模块。 <p class="p">在此用例中，qtimlvclassification 执行以下操作： </p><ol class="ol" type="a" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_p3v_2xl_vbc"><br>                                            <li class="li">加载模型的子模块。</li><br><br>                                            <li class="li">将结果生成为带有分类标签的视频帧。</li><br><br>                                            <li class="li">将它们发送至 qtivcomposer 的接收端。</li><br><br>                                        </ol><br></li><br><br>                                </ol> |
| [qtivcomposer](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qtivcomposer.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_nmc_lxl_vbc"><br>                                    <li class="li">在接收端上接收原始视频流和分类结果。 </li><br><br>                                    <li class="li">在其发送端上生成 GST 缓存，其内容由来自其接收端的视频流组成。</li><br><br>                                </ol> |
| **输出** | **输出** |
| [Waylandsink](https://docs.qualcomm.com/doc/80-70017-50SC/topic/waylandsink.html) | <ol class="ol" id="single-camera-stream-with-image-classification-and-display-with-mobilenet-v1__ol_cgt_mwl_vbc"><br>                                    <li class="li">在其接收端上接收视频</li><br><br>                                    <li class="li">将视频流提交到 Weston。 </li><br><br>                                    <li class="li">Weston 在本地显示设备上渲染视频流和为该场景生成的可能分类。</li><br><br>                                </ol> |

**上一级主题：** [Qualcomm Neural Processing SDK 用例](https://docs.qualcomm.com/doc/80-70017-50SC/topic/qualcomm-neural-processing-sdk-use-cases.html)

Last Published: Nov 11, 2025

[Previous Topic
Qualcomm Neural Processing SDK 用例](https://docs.qualcomm.com/bundle/publicresource/80-70017-50SC/topics/qualcomm-neural-processing-sdk-use-cases.md) [Next Topic
使用 Neural Processing SDK 进行图像分类和编码](https://docs.qualcomm.com/bundle/publicresource/80-70017-50SC/topics/single-camera-stream-with-image-classification-and-encode-with-mobilenet-v1.md)