自动裁剪人脸:简化你的数字人素材准备

ini复制代码def get_face_coordinates(image):   
    #cv读取的图片转为RGB格式,然后使用dlib进行人脸检测    
    rgb_image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)   
    detections = dlib.get_frontal_face_detector(rgb_image)    
    face = detections[0]   
    #计算并返回人脸中心点   
    center_x = (face.left() + face.right()) // 2    
    center_y = (face.top() + face.bottom()) // 2   
    return center_x, center_y

#读取图片
frame = cv2.imread(inputFile)
#进行人脸关键点检测,返回人脸中心点坐标
face_coords = get_face_coordinates(frame)
center_x, center_y = face_coords
#根据目标尺寸,计算出以人脸为中心点进行裁剪的起止坐标
crop_size = targetWH//2 
start_x = max(center_x - crop_size, 0)
start_y = max(center_y - crop_size, 0)
#把裁剪的关键点坐标信息记录,由于后期融合处理
crop_coordinates = {    "start_x": start_x,    "start_y": start_y,    "width": targetWH,    "height": targetWH}crop_coordinates_path = outputFamesFolder + '/crop_coordinates.json'with open(crop_coordinates_path, 'w') as file:    json.dump(crop_coordinates, file, indent=4)
#对原图进行裁剪
cropped_image = frame[start_y:start_y + targetWH, start_x:start_x + targetWH]
frame_filename = f"{outputFamesFolder}/frame_crop.png"
#保存裁剪后的图片cv2.imwrite(frame_filename, cropped_image)

ini复制代码#读取视频
video_capture = cv2.VideoCapture(inputFile)
#获取第一帧的人脸坐标
ret, first_frame = video_capture.read()
face_coords = get_face_coordinates(first_frame)
center_x, center_y = face_coords
crop_size = targetWH//2  
start_x = max(center_x - crop_size, 0)
start_y = max(center_y - crop_size, 0)
#根据第一帧的人脸坐标信息,逐帧进行裁剪
frame_number = 0
while video_capture.isOpened():   
    ret, frame = video_capture.read()    
    if not ret:        
       break   
    cropped_image = frame[start_y:start_y + targetWH, start_x:start_x + targetWH] 
    frame_filename = f"{outputFamesFolder}/frame_{frame_number:04d}.png"   
    cv2.imwrite(frame_filename, cropped_image)   
    frame_number += 1video_capture.release() 
    #最终裁剪后的视频
    pathoutput_video_path = outputFolder + f"/{name}_face_crop.mp4"
    # 提取视频的音频
    cmd = [    "ffmpeg",    "-i",inputFile,    '-f','wav',    '-ar','16000',    '-y',audio_output]
    out = subprocess.run(cmd,\        stdout=subprocess.PIPE,\        stderr=subprocess.STDOUT,\        text=True)
    #把裁剪人脸后的视频帧和音频无损合并为 最终的视频
    cmd =[    "ffmpeg",    "-i",frame_pattern,    "-i",audio_output,    "-c:v","libx264",    "-framerate", str(targetFps),    "-c:a","copy",    "-crf","0",    "-preset","veryslow",    "-pix_fmt", "yuv420p",    "-movflags","faststart",    "-shortest",    "-y",output_video_path]
    out = subprocess.run(cmd,\        stdout=subprocess.PIPE,\        stderr=subprocess.STDOUT,\        text=True)