Gen2 4K彩色相机运行MobileNetSSD模型

这个例子展示了如何在RGB预览流中运行mobileenetv2ssd模型,以及如何在预览中同时显示RGB预览和mobileenetv2ssd模型的元数据结果。预览大小设置为4K分辨率。

演示

设置

请运行以下命令来安装所需的依赖项

Warning

说明:此处安装的是第二代depthai库

python3 -m pip install --extra-index-url https://artifacts.luxonis.com/artifactory/luxonis-python-snapshot-local/ depthai==0.0.2.1+c9a19df719cb668e438d6eafd193cdf60a0d9354 numpy==1.19.5 opencv-python==4.5.1.48

有关更多信息,请参阅 Python API 安装指南

这个示例还需要 mobileenetsdd blob ( mobilenet.blob 文件 )才能工作——您可以从 这里 下载它。

源代码

可以在 GitHub 上找到。国内用户也可以在 gitee 上找到。

from pathlib import Path
import sys
import cv2
import depthai as dai
import numpy as np

# 首先获取模型
mobilenet_path = str((Path(__file__).parent / Path('models/mobilenet.blob')).resolve().absolute())
if len(sys.argv) > 1:
    mobilenet_path = sys.argv[1]

# 开始定义管道
pipeline = dai.Pipeline()

# 创建彩色相机流
cam_rgb = pipeline.createColorCamera()
cam_rgb.setPreviewSize(300, 300)    # 设置神经网络输入大小
# 设置彩色相机分辨率为4K
cam_rgb.setResolution(dai.ColorCameraProperties.SensorResolution.THE_4_K)
# 设置是否交错
cam_rgb.setInterleaved(False)
# 设置是否预览保持宽高比
cam_rgb.setPreviewKeepAspectRatio(False)

# 定义一个将基于源帧进行预测的神经网络
detection_nn = pipeline.createNeuralNetwork()
detection_nn.setBlobPath(mobilenet_path)
cam_rgb.preview.link(detection_nn.input)

# 创建输出流
xout_video = pipeline.createXLinkOut()
xout_video.setStreamName("video")
cam_rgb.video.link(xout_video.input)

xout_preview = pipeline.createXLinkOut()
xout_preview.setStreamName("preview")
cam_rgb.preview.link(xout_preview.input)

xout_nn = pipeline.createXLinkOut()
xout_nn.setStreamName("nn")
detection_nn.out.link(xout_nn.input)

# MobilenetSSD标签文本
texts = ["background", "aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow",
         "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"]

# 管道已定义,现在设备已连接到管道
with dai.Device(pipeline) as device:
    # 启动管道
    device.startPipeline()

    # 输出队列将用于从上面定义的输出中获取帧和nn数据
    q_video = device.getOutputQueue(name="video", maxSize=4, blocking=False)
    q_preview = device.getOutputQueue(name="preview", maxSize=4, blocking=False)
    q_nn = device.getOutputQueue(name="nn", maxSize=4, blocking=False)

    preview_frame = None
    video_frame = None
    bboxes = []
    labels = []
    confidences = []


    # nn数据(作为边界框的位置)在<0..1>范围内-需要使用图像的width/height对其进行归一化
    def frame_norm(frame, bbox):
        norm_vals = np.full(len(bbox), frame.shape[0])
        norm_vals[::2] = frame.shape[1]
        return (np.clip(np.array(bbox), 0, 1) * norm_vals).astype(int)


    def display_frame(name, frame, bboxes):
        for raw_bbox in bboxes:
            bbox = frame_norm(frame, raw_bbox)
            cv2.rectangle(frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]),
                        (255, 0, 0), 2)
        cv2.imshow(name, frame)


    while True:
        # 使用tryGet方法(非阻塞)而不是get方法来获取(阻塞),这将返回可用数据,否则返回None
        in_video = q_video.tryGet()
        in_preview = q_preview.tryGet()
        in_nn = q_nn.tryGet()

        if in_video is not None:
            # 如果摄像机的数据不为空,则将一维数据转换为HxWxC帧
            packetData = in_video.getData()
            w = in_video.getWidth()
            h = in_video.getHeight()
            yuv420p = packetData.reshape((h * 3 // 2, w))
            video_frame = cv2.cvtColor(yuv420p, cv2.COLOR_YUV2BGR_NV12)

        if in_preview is not None:
            shape = (3, in_preview.getHeight(), in_preview.getWidth())
            preview_frame = in_preview.getData().reshape(shape).transpose(1, 2, 0).astype(np.uint8)
            preview_frame = np.ascontiguousarray(preview_frame)

        if in_nn is not None:
            # 检测结果有7个数,最后一次检测后跟着-1位数,以后填充0
            bboxes = np.array(in_nn.getFirstLayerFp16())
            # 将一维数组转换为Nx7矩阵
            bboxes = bboxes.reshape((bboxes.size // 7, 7))
            # 筛选出置信度小于定义阈值的结果
            bboxes = bboxes[bboxes[:, 2] > 0.5]
            # 剪切bbox和标签
            labels = bboxes[:, 1].astype(int)
            confidences = bboxes[:, 2]
            bboxes = bboxes[:, 3:7]

        # 如果图像不为空,请在其上绘制边框并显示图像
        if video_frame is not None:
            for raw_bbox, label, conf in zip(bboxes, labels, confidences):
                bbox = frame_norm(video_frame, raw_bbox)
                cv2.rectangle(video_frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), (255, 0, 0), 2)
                cv2.putText(video_frame, texts[label], (bbox[0] + 10, bbox[1] + 20), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255)
                cv2.putText(video_frame, f"{int(conf * 100)}%", (bbox[0] + 10, bbox[1] + 40), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255)
            display_frame("video", video_frame, bboxes)

        if preview_frame is not None:
            for raw_bbox, label, conf in zip(bboxes, labels, confidences):
                bbox = frame_norm(preview_frame, raw_bbox)
                cv2.rectangle(preview_frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), (255, 0, 0), 2)
                cv2.putText(preview_frame, texts[label], (bbox[0] + 10, bbox[1] + 20), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255)
                cv2.putText(preview_frame, f"{int(conf * 100)}%", (bbox[0] + 10, bbox[1] + 40), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255)
            display_frame("preview", preview_frame, bboxes)

        if cv2.waitKey(1) == ord('q'):
            break

有疑问?

我们很乐意为您提供代码或其他问题的帮助。

我们的联系方式

OAK QQ群
QQ

OAK视觉人工智能俱乐部

群号:280844897

加微信邀请入群
WeChat

微信号:13951940532

关注微信公众号:OAK视觉人工智能开发
欢迎到淘宝选购
taobao
OAK中国官方淘宝店

还可以通过我们发布的视频和文章了解OAK