# Qualcomm Neural Processing SDK 用例

Source: [https://docs.qualcomm.com/doc/80-70018-50SC/topic/qualcomm-neural-processing-sdk-use-cases.html](https://docs.qualcomm.com/doc/80-70018-50SC/topic/qualcomm-neural-processing-sdk-use-cases.html)

Qualcomm Neural Processing SDK（原称为 Qualcomm Snapdragon Neural Processing Engine (SNPE)）用于运行深度神经网络进行推理。这些用例描述了使用不同 ML 模型的图像分类、目标检测和图像分割场景。

Note: Pipeline 中使用的深度学习容器 (DLC) 模型随 Qualcomm Neural Processing SDK 版本一同提供。

在运行用例之前，请完成 [GStreamer 命令行用例](https://docs.qualcomm.com/doc/80-70018-50SC/topic/gstreamer-application-use-cases.html)中提到的先决条件。

- **[使用 Neural Processing SDK 进行图像分类和显示](https://docs.qualcomm.com/doc/80-70018-50SC/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。
- **[使用 Neural Processing SDK 进行图像分类和编码](https://docs.qualcomm.com/doc/80-70018-50SC/topic/single-camera-stream-with-image-classification-and-encode-with-mobilenet-v1.html)**  

这些用例使用 InceptionV3 图像分类模型和 Qualcomm Neural Processing SDK，对单个摄像头流执行场景分类，并叠加或合成分类标签。然后对流进行编码。
- **[使用 Neural Processing SDK 进行目标检测和显示](https://docs.qualcomm.com/doc/80-70018-50SC/topic/single-camera-stream-with-object-detection-and-display-with-mobilenet-v2-ssd.html)**  

这些用例使用 yolonas.dlc 目标检测模型和 Qualcomm Neural Processing SDK 对单个摄像头流执行目标检测。该用例对检测到的对象上叠加或合成边界框，然后显示结果。
- **[使用 Neural Processing SDK 进行目标检测和编码](https://docs.qualcomm.com/doc/80-70018-50SC/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 对单个摄像头流执行目标识别。该用例对检测到的对象上叠加或合成边界框，然后将此流编码为 H.264 码流。
- **[使用 Neural Processing SDK 进行图像分割和显示](https://docs.qualcomm.com/doc/80-70018-50SC/topic/single-camera-stream-with-image-segmentation-and-display-with-deeplabv3-quantized.html)**  

该用例使用 DeepLab v3 模型和 Qualcomm Neural Processing SDK runtime 对单个摄像头流的场景进行语义分割识别。该用例使用 qtivcomposer 将语义和视频流合成在一起，然后显示结果。
- **[使用 Neural Processing SDK 进行图像分割和编码](https://docs.qualcomm.com/doc/80-70018-50SC/topic/single-camera-stream-with-image-segmentation-and-encode-with-deeplabv3-quantized.html)**  

该用例使用 DeepLab v3 模型和 Qualcomm Neural Processing SDK runtime。该用例将语义分割与原始视频流进行合成，对该流进行编码，然后将其复用到一个 MP4 容器中。

**Parent Topic:** [运行机器学习用例](https://docs.qualcomm.com/doc/80-70018-50SC/topic/machine-learning-use-cases.html)

**Related Resources**  

- [配置 Qualcomm GStreamer 插件](https://docs.qualcomm.com/doc/80-70018-50SC/topic/qim-sdk-plugins.html)

Last Published: Nov 12, 2025

[Previous Topic
从摄像头到 RTSP 的单流，带 ML 检测](https://docs.qualcomm.com/bundle/publicresource/80-70018-50SC/topics/single-stream-from-camera-to-rtsp-with-ml-detection.md) [Next Topic
使用 Neural Processing SDK 进行图像分类和显示](https://docs.qualcomm.com/bundle/publicresource/80-70018-50SC/topics/single-camera-stream-with-image-classification-and-display-with-mobilenet-v1.md)