97 lines
2.4 KiB
Markdown
97 lines
2.4 KiB
Markdown
|
# Camera Terminal[[English](./README.md)]
|
|||
|
|
|||
|
该示例的输入图片来自摄像头,输出结果打印在终端。该示例演示了以下模型接口在实际项目中的使用情况。
|
|||
|
|
|||
|
- [HumanFaceDetectMSR01](https://github.com/espressif/esp-dl/blob/master/include/model_zoo/human_face_detect_msr01.hpp)
|
|||
|
- [HumanFaceDetectMNP01](https://github.com/espressif/esp-dl/blob/master/include/model_zoo/human_face_detect_mnp01.hpp)
|
|||
|
- [CatFaceDetectMN03](https://github.com/espressif/esp-dl/blob/master/include/model_zoo/cat_face_detect_mn03.hpp)
|
|||
|
|
|||
|
|
|||
|
|
|||
|
## 硬件要求:
|
|||
|
|
|||
|
芯片: ESP32,ESP32-S2,ESP32-S3
|
|||
|
|
|||
|
摄像头: OV2640,OV3660
|
|||
|
|
|||
|
对于新手,我们为您推荐开发套件:[ESP-EYE](https://www.espressif.com/zh-hans/products/devkits/esp-eye/overview),ESP-S3-EYE,[ESP32-WROVER-KIT](https://docs.espressif.com/projects/esp-idf/en/latest/esp32/hw-reference/esp32/get-started-wrover-kit-v3.html)
|
|||
|
|
|||
|
|
|||
|
|
|||
|
## 示例演示:
|
|||
|
|
|||
|
如果您完成了硬件准备并已正确安装好摄像头,那我们可以开始软件演示了。
|
|||
|
|
|||
|
1. 打开终端,并进入当前示例所在目录路经:
|
|||
|
|
|||
|
```shell
|
|||
|
cd ~/esp-who/examples/camera_terminal
|
|||
|
```
|
|||
|
|
|||
|
|
|||
|
|
|||
|
2. 设定目标芯片:
|
|||
|
|
|||
|
```shell
|
|||
|
idf.py set-target [SoC]
|
|||
|
```
|
|||
|
|
|||
|
将 [SoC] 替换为您的目标芯片,如 esp32、 esp32s2、esp32s3。
|
|||
|
|
|||
|
|
|||
|
|
|||
|
3. 打开 menuconfig,设置参数:
|
|||
|
|
|||
|
```shell
|
|||
|
idf.py menuconfig
|
|||
|
```
|
|||
|
|
|||
|
1. 选用自定义分区:
|
|||
|
|
|||
|
![](../../img/menuconfig_custom_partition_table.png)
|
|||
|
|
|||
|
|
|||
|
|
|||
|
2. 选择正确的摄像头管脚配置,下图以 ESP-EYE 为例:
|
|||
|
|
|||
|
![](../../img/menuconfig_select_camera_pinout.png)
|
|||
|
|
|||
|
如果您使用的是非官方支持的开发组件,请在 Custom Camera Pinout 中配置正确管脚信息。
|
|||
|
|
|||
|
|
|||
|
|
|||
|
3. 选择您想要运行的模型接口:
|
|||
|
|
|||
|
![](./img/menuconfig_dl_config.png)
|
|||
|
|
|||
|
|
|||
|
|
|||
|
4. 打开外部 RAM:
|
|||
|
|
|||
|
![](../../img/menuconfig_support_external_ram.png)
|
|||
|
|
|||
|
|
|||
|
|
|||
|
注意:
|
|||
|
|
|||
|
- 每次设定芯片之后,都需要重新设定以上参数。
|
|||
|
|
|||
|
- 除了以上参数,用户也可以根据应用场景配置 CPU、Flash等的其他参数。
|
|||
|
|
|||
|
|
|||
|
|
|||
|
4. 烧录程序,运行 IDF 监视器获取实时结果:
|
|||
|
|
|||
|
```shell
|
|||
|
idf.py flash monitor
|
|||
|
```
|
|||
|
|
|||
|
以步骤 3 中的模型选择为例,打印结果如下所示:
|
|||
|
|
|||
|
![](./img/result_on_terminal.png)
|
|||
|
|
|||
|
打印结果中包括时间消耗和检测结果。
|
|||
|
|
|||
|
|
|||
|
|