查看“JetCam”的源代码

== '''JetCam''' ==
'''JetCam是一个容易上手、适用于NVIDIA Jetson的Python相机。'''

* 使用Jetson的[https://developer.download.nvidia.cn/embedded/L4T/r32_Release_v1.0/Docs/Accelerated_GStreamer_User_Guide.pdf?uIzwdFeQNE8N-vV776ZCUUEbiJxYagieFEqUoYFM9XSf9tbslxWqFKnVHu8erbZZS20A7ADAIgmSQJvXZTb0LkuGl9GoD5HJz4263HcmYWZW0t2OeFSJKZOfuWZ-lF51Pva2DSDtu2QPs-junm7BhMB_9AMQRwExuDb5zIhf_o8PIbA4KKo Accelerated GStreamer插件]可与各种USB和CSI摄像机配合使用

* 使用image = camera.read（）可轻松将图像读取为numpy数组

* 设置相机以running = True将回调附加到新框架

JetCam让在Python中创建AI项目的原型变得容易，尤其是在[https://github.com/NVIDIA-AI-IOT/jetcard JetCard]中安装的Jupyter Lab编程环境中。

如果您发现问题，请[https://github.com/NVIDIA-AI-IOT/jetcam/issues 告诉我们]！

== '''设定''' ==
# git clone https://github.com/NVIDIA-AI-IOT/jetcam
# cd jetcam
# sudo python3 setup.py install

*JetCam已针对配置了[https://github.com/NVIDIA-AI-IOT/jetcard JetCard]设置的系统进行了测试。不同的系统配置可能需要其他步骤。

== '''用法''' ==
下面我们列举一些用法示例。您可以在[https://github.com/NVIDIA-AI-IOT/jetcam/tree/master/notebooks notebooks]中找到更多内容。

== '''创建CSI摄像机''' ==
* 调用CSICamera函数以使用兼容的CSI摄像机
* 其中capture_width、capture_height和capture_fps将控制获取图像的捕获形状和速率
* width和height控制该read函数返回图像的最终输出形状。

# from jetcam.csi_camera import CSICamera
# camera = CSICamera(width=224, height=224, capture_width=1080, capture_height=720, capture_fps=30)

== '''创建USB相机''' ==
* 调用USBCamera函数以使用兼容的USB相机，应用与CSICamera 相同的参数，以及指示设备索引的参数capture_device。您可以通过调用ls /dev/video*来检查设备索引。

# from jetcam.usb_camera import USBCamera
# camera= USBCamera（capture_device=1）

== '''读取''' ==
*调用read()函数以读取最新图像numpy.ndarray的数据类型np.uint8和形状(224, 224, 3)。颜色格式为BGR8。

# image = camera.read（）

*read函数还会更新相机的内部value属性。

# camera.read（）
# image = camera.value

== '''回调''' ==
*您也可以将相机设置为running = True，这将产生一个从相机获取图像的线程，会自动更新相机的value属性。
*您可以使用[https://traitlets.readthedocs.io/en/stable/api.html#callbacks-when-trait-attributes-change traitlets库]将回调函数附加到该值，这将使用新的相机值和旧的相机值调用回调。

#  camera.running = True
# def callback(change):
#     new_image = change['new']
#     #   做一些处理......
# camera.observe(callback, names='value')

== '''摄像头''' ==
=== '''CSI相机''' ===
*以下相机与[https://github.com/NVIDIA-AI-IOT/jetcam/blob/master/jetcam/csi_camera.py CSICamera类]一起使用，可以按照[https://github.com/NVIDIA-AI-IOT/jetcam/blob/master/notebooks/csi_camera/csi_camera.ipynb notebook]的例子来尝试它们。
{| class="wikitable"
|-
! 型号 !! 红外 !! 视角 !! 分辨率 !! 价格
|-
| [http://www.waveshare.net/shop/RPi-Camera-V2.htm Raspberry Pi Camera V2] ||  || 62.2 || 3280x2464 || 143 RMB
|-
| [http://www.waveshare.net/shop/RPi-NoIR-Camera-V2.htm Raspberry Pi Camera V2 (NOIR)] ||  √ || 62.2 || 3280x2464 || 148 RMB
|-
| [https://www.robotshop.com/en/arducam-8mp-sony-imx219-camera-module-cs-lens-2718-raspberry-pi.html?gclid=EAIaIQobChMIzMKg38bD4QIVrR6tBh3UoAdjEAYYCSABEgLg-_D_BwE Arducam IMX219 CS lens mount] ||  ||  || 3280x2464 || 455 RMB
|-
| [https://www.robotshop.com/en/arducam-8mp-sony-imx219-camera-module-m12-lens-ls40136-raspberry-pi.html Arducam IMX219 M12 lens mount] ||  ||  || 3280x2464 || 420 RMB
|-
| [https://leopardimaging.com/product/nvidia-jetson-cameras/nvidia_nano_mipi_camera_kits/li-imx219-mipi-ff-nano/ LI-IMX219-MIPI-FF-NANO] ||  ||  || 3280x2464 || 203 RMB
|-
| [http://www.waveshare.net/shop/IMX219-77-Camera.htm WaveShare IMX219-77] ||  || 77 || 3280x2464 || 98 RMB
|-
| [http://www.waveshare.net/shop/IMX219-77IR-Camera.htm WaveShare IMX219-77IR] ||  √ || 77 || 3280x2464 || 113 RMB
|-
| [http://www.waveshare.net/shop/IMX219-120-Camera.htm WaveShare IMX219-120] ||  || 120 || 3280x2464 || 108 RMB
|-
| [http://www.waveshare.net/shop/IMX219-160-Camera.htm WaveShare IMX219-160] ||  || 160 || 3280x2464 || 125 RMB
|-
| [http://www.waveshare.net/shop/IMX219-160IR-Camera.htm WaveShare IMX219-160IR] ||   √ || 160 || 3280x2464 || 138 RMB
|-
| [http://www.waveshare.net/shop/IMX219-200-Camera.htm WaveShare IMX219-200] ||  || 200 || 3280x2464 || 148 RMB
|}

=== '''USB相机''' ===
*以下相机与[https://github.com/NVIDIA-AI-IOT/jetcam/blob/master/jetcam/usb_camera.py USBCamera类]一起使用，可以按照[https://github.com/NVIDIA-AI-IOT/jetcam/blob/master/notebooks/usb_camera/usb_camera.ipynb notebook]的例子来尝试它们。
{| class="wikitable"
|-
! 型号 !! 红外 !! 视角 !! 分辨率 !! 价格
|-
| [https://www.amazon.com/Logitech-Widescreen-designed-Calling-Recording/dp/B004FHO5Y6 Logitech C270] ||  || 60 || 1280x720 || 126 RMB
|}

== '''更多资讯如下''' ==
* [https://github.com/NVIDIA-AI-IOT/jetbot JetBot]——基于NVIDIA Jetson Nano的教育型AI机器人

* [https://github.com/NVIDIA-AI-IOT/jetracer JetRacer]——使用NVIDIA Jetson Nano的教育性AI赛车

* [https://github.com/NVIDIA-AI-IOT/jetcard JetCard]——用于使用NVIDIA Jetson Nano进行Web编程AI项目的SD卡映像

* [https://github.com/NVIDIA-AI-IOT/torch2trt torch2trt]——易于使用的PyTorch到TensorRT转换器