Python中OpenCV处理网络摄像头视频转灰度后无法播放问题排查

我想用OpenCV把网络摄像头的RTSP流转换成灰度图并保存为视频文件，但生成的detect.avi无法播放。运行代码时出现cv::icvExtractPattern断言失败错误，添加openh264-1.8.0-win64.dll后又出现libopenh264初始化失败等一系列错误，代码及报错信息如下：

原代码 #

rtsp_path = f"rtsp://{userid}:{pwd}@{ip_address}:{rtsp_port}/ipcam_h264.sdp"
video = cv2.VideoCapture(rtsp_path )
savepath = "detect.avi"
frame_Rate = video.get(cv2.CAP_PROP_FPS)
height, width, dim = frame.shape
rec = cv2.VideoWriter(savepath, cv2.VideoWriter_fourcc(*'DIVX'), frame_Rate, (int(width), int(height)))

while (video.isOpened()):
    ret, frame = video.read()
    if frame is None:
        break

    img = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    rec.write(img)

    if cv2.waitKey(30) & 0xFF == ord('q'):
        video.release()
        rec.release()
        cv2.destroyAllWindows()
        break

cv2.destroyAllWindows()
video.release()
rec.release()

原始报错信息 #

OpenCV(4.6.0) D:\a\opencv-python\opencv-python\opencv\modules\videoio\src\cap_images.cpp:267: error: (-215:Assertion failed) number < max_number in function 'cv::icvExtractPattern'

添加openh264-1.8.0-win64.dll后的报错信息 #

OpenH264 Video Codec provided by Cisco Systems, Inc.

[libopenh264 @ 000001b9e01c4700] [OpenH264] this = 0x000001B9DA5EC4D0, Error:MaxSpatialBitrate (288000000) should be larger than SpatialBitrate (536870911), considering it as error setting
[libopenh264 @ 000001b9e01c4700] [OpenH264] this = 0x000001B9DA5EC4D0, Error:WelsInitEncoderExt(), ParamValidationExt failed return 16.
[libopenh264 @ 000001b9e01c4700] [OpenH264] this = 0x000001B9DA5EC4D0, Error:CWelsH264SVCEncoder::Initialize(), WelsInitEncoderExt failed.
[libopenh264 @ 000001b9e01c4700] Initialize failed
[ERROR:0@15.734] global /build/opencv/modules/videoio/src/cap_ffmpeg_impl.hpp (2985) open Could not open codec libopenh264, error: Unspecified error (-1313558101)
[ERROR:0@15.734] global /build/opencv/modules/videoio/src/cap_ffmpeg_impl.hpp (3002) open VIDEOIO/FFMPEG: Failed to initialize VideoWriter
[ERROR:0@24.573] global D:\a\opencv-python\opencv-python\opencv\modules\videoio\src\cap.cpp (597) cv::VideoWriter::open VIDEOIO(CV_IMAGES): raised OpenCV exception:

OpenCV(4.6.0) D:\a\opencv-python\opencv-python\opencv\modules\videoio\src\cap_images.cpp:267: error: (-215:Assertion failed) number < max_number in function 'cv::icvExtractPattern'

问题根源分析 #

1. 空指针访问错误：代码在调用video.read()前就直接访问frame.shape，此时frame未初始化，是None，直接触发断言失败。
2. 通道格式不匹配：转换后的灰度图是单通道，但VideoWriter初始化时默认按3通道配置，写入单帧时格式不兼容。
3. 编码库冲突：原代码使用DIVX编码，额外添加OpenH264库会干扰默认编码逻辑，导致码率参数冲突引发初始化失败。

修正后的代码 #

import cv2

# 替换为你的RTSP参数
userid = "your_user"
pwd = "your_pwd"
ip_address = "your_ip"
rtsp_port = "554"

rtsp_path = f"rtsp://{userid}:{pwd}@{ip_address}:{rtsp_port}/ipcam_h264.sdp"
video = cv2.VideoCapture(rtsp_path)
savepath = "detect.avi"

# 先读取一帧确认流有效
ret, frame = video.read()
if not ret or frame is None:
    print("无法读取RTSP流")
    exit()

frame_Rate = video.get(cv2.CAP_PROP_FPS)
height, width = frame.shape[:2]
# 指定isColor=False，匹配灰度图单通道格式
rec = cv2.VideoWriter(savepath, cv2.VideoWriter_fourcc(*'DIVX'), frame_Rate, (width, height), isColor=False)

while video.isOpened():
    ret, frame = video.read()
    if not ret or frame is None:
        break

    img = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    rec.write(img)

    if cv2.waitKey(30) & 0xFF == ord('q'):
        break

# 统一释放资源
video.release()
rec.release()
cv2.destroyAllWindows()

关键修正点 #

• 先读取有效帧再获取图像尺寸，避免空指针错误。
• 初始化VideoWriter时添加isColor=False，适配灰度图单通道格式。
• 移除额外添加的OpenH264库，避免编码逻辑冲突。
• 优化资源释放逻辑，统一放在循环外执行。

内容的提问来源于stack exchange，提问作者YuK