42 lines
1.5 KiB
Markdown
42 lines
1.5 KiB
Markdown
|
# Face Detection and Recognition with ESP-EYE
|
|||
|
|
|
|||
|
|
This example is supposed to demonstrate the elementry function of the board [ESP-EYE](), which is based on ESP32 chip and intergrated with camera, led, and digital microphone modules[detection_with_command_line example](./detection_with_command_line), but displaies the image stream through http server.
|
|||
|
|
|
|||
|
|
## General Information
|
|||
|
|
|
|||
|
|
用户可以通过连接esp32建立的softap，在局域网内访问其特定的网址，不管是手机端还是电脑端，都可以在网页上看到图像拍摄和检测的效果。
|
|||
|
|
|
|||
|
|
## 安装及使用
|
|||
|
|
|
|||
|
|
- 通过`make defconfig`使用默认配置
|
|||
|
|
- 通过`make menuconfig`设置Wi-Fi的用户名和密码
|
|||
|
|
- 烧入程序`make flash`
|
|||
|
|
- 找到一个手机或者电脑，连接上建立的softap
|
|||
|
|
- 在浏览器中输入`192.168.4.1/face_stream`
|
|||
|
|
- Please enjoy it
|
|||
|
|
|
|||
|
|
## 图像格式及尺寸
|
|||
|
|
|
|||
|
|
我们建议使用默认的320x240的像素大小，因为更大的像素会带来更多的计算。
|
|||
|
|
|
|||
|
|
输入输出的图像格式都使用jpeg，比一般的RGB小很多，适合网络的传输。
|
|||
|
|
|
|||
|
|
## 配置(Configuration)
|
|||
|
|
|
|||
|
|
- 可增加自己定制的handler，例如本例子使用的就是
|
|||
|
|
```
|
|||
|
|
httpd_uri_t _face_stream_handler = {
|
|||
|
|
.uri = "/face_stream",
|
|||
|
|
.method = HTTP_GET,
|
|||
|
|
.handler = facenet_stream_handler,
|
|||
|
|
.user_ctx = NULL
|
|||
|
|
};
|
|||
|
|
```
|
|||
|
|
在初始化过程中进行注册:
|
|||
|
|
`httpd_register_uri_handler(camera_httpd, &_face_stream_handler);`
|
|||
|
|
|
|||
|
|
- 可增加camera的缓存队列，提高并行的速度
|
|||
|
|
`camera_config_t config.fb_count = 2`
|
|||
|
|
|
|||
|
|
|