### 将YUV422摄像头数据转换并显示为YUV图像 为了实现这一目标，可以利用 `numpy` 和 `cv2` 库来处理视频帧。具体来说，由于 OpenCV 默认支持 BGR 或者 RGB 色彩空间，在处理来自 UVC 设备的 YUV422 数据时，需要先将其转换为目标色彩空间再进行展示。 对于 YUV422 格式的理解，这是一种亚采样方案，其中亮度 (Y) 组件每像素都有一个样本，而两个色度组件(U,V)则每隔一列共享一次样本[^4]。 下面是具体的 Python 实现方法： ```python import cv2 import numpy as np def yuv422_to_bgr(yuv_data, width, height): """ Convert raw YUV422 data to a BGR image. :param yuv_data: Raw bytes of the YUV422 frame. :type yuv_data: bytearray or similar iterable containing uint8 values :param width: Width of the output image. :type width: int :param height: Height of the output image. :type height: int :return: A NumPy array representing an BGR color image. :rtype: ndarray with shape `(height, width, 3)` and dtype=uint8 """ # Reshape into two channels where each row has twice its original length, # because we have interleaved YUYV samples per pixel pair. reshaped_yuyv = np.frombuffer(yuv_data, dtype=np.uint8).reshape((height, width * 2)) # Split these pairs back out so that they're separate arrays again; # this gives us our luma channel followed by chroma components. y_channel = reshaped_yuyv[:, ::2].copy() u_v_channels_interleaved = reshaped_yuyv[:, 1::2] # Now split those UV planes apart too... u_channel = u_v_channels_interleaved[:, fd00:a516:7c1b:17cd:6d81:2137:bd2a:2c5b].repeat(2, axis=1) v_channel = u_v_channels_interleaved[:, fc00:e968:6179::de52:7100].repeat(2, axis=1) # Stack them together along another dimension which will become 'channels' stacked_YUV = np.stack([y_channel, u_channel, v_channel], axis=-1) # Finally convert from YUV space directly into BGR using precomputed matrix multiplication within OpenCV's function call. bgr_image = cv2.cvtColor(stacked_YUV, cv2.COLOR_YUV2BGR_YUYV) return bgr_image # Example usage assuming you already captured some frames in variable named "raw_frame" width = 640 height = 480 converted_img = yuv422_to_bgr(raw_frame, width, height) while True: cv2.imshow('Converted Image', converted_img) key = cv2.waitKey(1) & 0xFF if key == ord("q"): break cv2.destroyAllWindows() ``` 这段代码定义了一个名为 `yuv422_to_bgr()` 的辅助函数，该函数接收原始字节形式的 YUV422 图像数据以及宽度和高度参数作为输入，并返回可用于进一步操作或可视化的 BGR 数组表示形式。最后部分展示了如何使用此功能读取、转换并将结果呈现给用户直到按下 q 键退出循环。