MacBook M1运行OpenCV采集关键点后Jupyter内核崩溃如何解决

使用设备为MacBook Air M1，开发行为检测程序时，为训练模型采集关键点运行了如下代码：

关键点采集代码 #

cap = cv2.VideoCapture(0)

# Set mediapipe model 
with mp_holistic.Holistic(min_detection_confidence=0.5, min_tracking_confidence=0.5) as holistic:
    
    for action in actions:
        # Loop through sequences aka videos
        for sequence in range(no_sequences):
            # Loop through video length aka sequence length, each frame, 30 frames of keypoints for each video
            for frame_num in range(sequence_length):

                # Read feed
                ret, frame = cap.read()

                # Make detections
                image, results = mediapipe_detection(frame, holistic)
                # print the results

                # Draw landmarks
                draw_styled_landmarks(image, results)
                
                # NEW Apply wait logic
                if frame_num == 0: 
                    cv2.putText(image, 'STARTING COLLECTION', (120,200), 
                               cv2.FONT_HERSHEY_SIMPLEX, 1, (0,255, 0), 4, cv2.LINE_AA)
                    cv2.putText(image, 'Collecting frames for {} Video Number {}'.format(action, sequence), (15,12), 
                               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)
                    # Show to screen
                    cv2.imshow('OpenCV Feed', image)
                    cv2.waitKey(1200)
                else: 
                    cv2.putText(image, 'Collecting frames for {} Video Number {}'.format(action, sequence), (15,12), 
                               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)
                    # Show to screen
                    cv2.imshow('OpenCV Feed', image)
                
                # NEW Export keypoints
                keypoints = extract_keypoints(results)
                npy_path = os.path.join(DATA_PATH, action, str(sequence), str(frame_num))
                np.save(npy_path, keypoints)

                # Break gracefully
                if cv2.waitKey(10) & 0xFF == ord('q'):
                    break
                    
    cap.release()
    cv2.destroyAllWindows()

报错现象 #

• 运行上述代码完成数据录制后，OpenCV feed窗口持续加载无响应，Mac出现彩虹转圈光标
• 运行如下模型训练代码时Python意外退出，Jupyter Notebook内核崩溃，提示Python quit unexpectedly. The kernel appears to have died. It will restart automatically.

# Training Model
# Preprocess Data
from sklearn.model_selection import train_test_split

from tensorflow.keras.utils import to_categorical 

尝试操作及报错 #

尝试通过新建conda环境解决OpenCV相关问题，执行命令conda create --name myenv opencv-env python=3.6后运行失败，终端返回如下内容：

Collecting package metadata (current_repodata.json): done
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed

PackagesNotFoundError: The following packages are not available from current channels:

  - opencv-env

Current channels:

  - https://repo.anaconda.com/pkgs/main/osx-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/r/osx-64
  - https://repo.anaconda.com/pkgs/r/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.

• 修正conda创建环境的错误命令：之前的命令混淆了虚拟环境名和依赖包名，opencv-env是要创建的环境名，不是包名，且M1芯片不支持Python3.6，最低适配3.7及以上版本，推荐用兼容性更好的3.9版本，正确的创建命令为：

conda create --name opencv-env python=3.9

• 环境创建完成后执行激活命令：

conda activate opencv-env

• 安装适配M1芯片的依赖包，优先选用conda-forge源的arm64适配版本，避免默认源的x86架构包不兼容问题：

conda install -c conda-forge opencv-python mediapipe tensorflow-macos tensorflow-metal scikit-learn jupyter

• 修复采集代码的逻辑bug：原有代码中按q触发的break只会跳出最内层的帧循环，cap.release()和cv2.destroyAllWindows()只有所有数据采集完成才会执行，中途退出会导致窗口进程残留无响应，修改退出逻辑：

# Break gracefully
if cv2.waitKey(10) & 0xFF == ord('q'):
    cap.release()
    cv2.destroyAllWindows()
    exit()

• 后续启动Jupyter Notebook前需先激活对应的conda环境，再执行jupyter notebook启动服务，保证内核调用的是适配环境的依赖包。

内容的提问来源于stack exchange，提问作者teeara