# qtimlvconverter

Source: [https://docs.qualcomm.com/doc/80-70018-50SC/topic/qtimlvconverter.html](https://docs.qualcomm.com/doc/80-70018-50SC/topic/qtimlvconverter.html)

qtimlvconverter 插件将传入的视频缓存转换为神经网络张量，同时在此过程中执行必要的格式转换和调整大小。为了实现这些操作，该插件使用 GPU 硬件和 ION/DMA 分配的缓存。

关于浮点张量，需要使用平均值和西格玛属性来使用公式：变换来自视频源的 8 位无符号整数镜像

(channel - mean) × sigma 其中通道是颜色转换输入的红色、绿色、蓝色或 Alpha 通道的值。

例如，下表列出了影响图像通道变换的平均值和西格玛属性：

| 属性 | 带通道变换的 RGB 图像 |
| --- | --- |
| R=255，G=127，B=0 | R=(255 - 128.0) x 0.75, G=(127 - 156.0)<br>                                x 0.34, B=(0 - 124.0) x 0.02 |
| 平均值 =“&lt;128.0, 156.0, 124.0, 58.0&gt;” | R=(255 - 128.0) x 0.75, G=(127 - 156.0)<br>                                x 0.34, B=(0 - 124.0) x 0.02 |
| sigma="&lt;0.75, 0.34, 0.02, 0.07&gt;" | R=(255 - 128.0) x 0.75, G=(127 - 156.0)<br>                                x 0.34, B=(0 - 124.0) x 0.02 |
|  |  |
|  |  |

subpixel-layout 属性指定输出张量中图像像素的排列。

该插件使用 Qualcomm IB2C 库进行所有转换操作。该库与 API 封装在自定义 GstGlesVideoConverter 抽象层中，用于创建、配置和处理传入和传出缓存。

GstMLBufferPool，一个自定义缓存类，可以执行以下操作：

- 分配输出缓存。
- 通过 IOCTL 命令向内核分配 ION 缓存。

Figure : 使用 qtimlvconverter 的 Gstbuffer 工作流
            
            ![](data:image/png;base64,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)

## 继承链

[GObject](https://docs.gtk.org/gobject/) → [GstObject](https://gstreamer.freedesktop.org/documentation/gstreamer/gstobject.html?gi-language=c) → [GstElement](https://gstreamer.freedesktop.org/documentation/gstreamer/gstelement.html?gi-language=c) → [GstBaseTransform](https://gstreamer.freedesktop.org/documentation/base/gstbasetransform.html?gi-language=c) → GstMLVideoConverter

下表提供了有关 qtimlvconverter 的端口模板和元素属性的信息。有关用例信息，请参见 [LiteRT 用例](https://docs.qualcomm.com/doc/80-70018-50SC/topic/tensorflow-lite-use-cases.html)和 [Qualcomm Neural Processing SDK 用例](https://docs.qualcomm.com/doc/80-70018-50SC/topic/qualcomm-neural-processing-sdk-use-cases.html)。

## 端口配置

| 端口名称 | 功能 | 功能 | 功能 |
| --- | --- | --- | --- |
| SINK 模板：'sink'<br><br><br>                                <br>*可用性：* 始终可用<br><br><br>                                <br>*方向：* 接收 | video/x-raw | format： | { (string)RGBA, (string)BGRA, (string)ABGR, (string)ARGB, (string)RGBx, (string)BGRx, (string)xRGB, (string)xBGR, (string)BGR, (string)RGB, (string)GRAY8, (string)NV12, (string)NV21, (string)YUY2, (string)UYVY } |
| SINK 模板：'sink'<br><br><br>                                <br>*可用性：* 始终可用<br><br><br>                                <br>*方向：* 接收 | video/x-raw (memory:GBM) | format： | { (string)RGBA, (string)BGRA, (string)ABGR, (string)ARGB, (string)RGBx, (string)BGRx, (string)xRGB, (string)xBGR, (string)BGR, (string)RGB, (string)GRAY8, (string)NV12, (string)NV21, (string)YUY2, (string)UYVY } |
| SRC 模板：'src'<br><br><br>                                <br>*可用性：* 始终可用<br><br><br>                                <br>*方向：* 发送 | neural-network/tensors | type： | { (string)UINT8, (string)INT32, (string)FLOAT16, (string)FLOAT32 } |
|  |  |  |  |

## 元素配置

Table : qtimlvconverter 的元素属性

| 属性 | 说明 |
| --- | --- |
| name | 对象名称<ul class="ul" id="qtimlvconverter__ul_rqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li">字符串。默认："mlvideoconverter0"</li><br><br>                                </ul> |
| parent | 对象的父级<ul class="ul" id="qtimlvconverter__ul_sqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li">"GstObject" 类型的对象</li><br><br>                                </ul> |
| qos | 处理服务质量事件。<ul class="ul" id="qtimlvconverter__ul_tqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li">Boolean。默认值：false</li><br><br>                                </ul> |
| subpixel-layout | 输出张量中图像像素的排列。<ul class="ul" id="qtimlvconverter__ul_uqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li"><br>                                        <p class="p">枚举 “GstMLVideoPixelLayout” 默认：0，“regular”</p><ul class="ul" id="qtimlvconverter__ul_c1t_lh4_myb"><br>                                            <li class="li"> (0)：regular - 常规子像素布局。例如，RGB、RGBA 或 RGBx。</li><br><br>                                            <li class="li"> (1)：reverse – 反向子像素布局。例如，BGR、BGRA 或 BGRx。</li><br><br>                                        </ul><br><br>                                    </li><br><br>                                </ul> |
| mean | 通道均值 FLOAT 张量的减法值 ('&lt;R, G, B&gt;', '&lt;R, G, B, A&gt;', '&lt;G&gt;')<ul class="ul" id="qtimlvconverter__ul_vqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li">"gdouble" 类型 GValue 的 GstValueArray</li><br><br>                                </ul> |
| sigma | FLOAT 张量的通道除数值（'&lt;R, G, B&gt;', '&lt;R, G, B, A&gt;', '&lt;G&gt;'）<ul class="ul" id="qtimlvconverter__ul_wqy_kh4_myb"><br>                                    <li class="li">标志：可读、可写</li><br><br>                                    <li class="li">"gdouble" 类型 GValue 的 GstValueArray</li><br><br>                                </ul> |
| engine | 选择 Qualcomm® Computer Vision SDK 或 GLES 作为 qtimlvconverter 的后端<ul class="ul" id="qtimlvconverter__ul_edm_ct3_tzb"><br>                                    <li class="li">值可以是 "fcv" 或 "gles"</li><br><br>                                    <li class="li"> 对于非 GPU 设备，默认值为 fcv。对于 GPU 设备，默认值为 gles</li><br><br>                                </ul><br><br>Note: 目前不支持 Qualcomm Computer Vision SDK（fcv）引擎。 |

**Parent Topic:** [配置 ML 插件](https://docs.qualcomm.com/doc/80-70018-50SC/topic/inferencing-plugins.html)

Last Published: Nov 12, 2025

[Previous Topic
配置 ML 插件](https://docs.qualcomm.com/bundle/publicresource/80-70018-50SC/topics/inferencing-plugins.md) [Next Topic
qtimlaconverter](https://docs.qualcomm.com/bundle/publicresource/80-70018-50SC/topics/qtimlaconverter.md)