Mediapipe,将地标分配给顶点?

Mediapipe, assign the landmarks to the vertices?

Here 是 mediapipe 用于其面部网格模型的 fbx 格式的面部。它有 468 个顶点。 Here 是指数的可视化。

Here是mediapipes人脸网格模型的描述。它输出地标位置。

我怎么知道哪个地标属于哪个顶点? 例如在搅拌机中。当我导入fbx人脸时,如何获得与mediapipe人脸网格模型的地标相同的索引?

似乎带有 fbx 模型的搅拌器中的索引与 mediapipe 面网格解决方案提供的索引相同。这些索引与 mediapipe 中的相同 canonical face model uv visualization. This answer 提供了通过其索引获取地标的示例。

需要启用 Developer Extras。在编辑模式下,选项显示在 Viewport Overlays > Developer > Indices 下,如下所示,以在搅拌机中获取索引。获取索引的替代选项可以是 found here.

我在下面展示了一个示例,其中左眼界标索引出现在规范的面部网格 uv 可视化中。

指数可视化

代码

代码基于,https://google.github.io/mediapipe/solutions/face_mesh.html

import cv2
import mediapipe as mp
mp_drawing = mp.solutions.drawing_utils
mp_drawing_styles = mp.solutions.drawing_styles
mp_face_mesh = mp.solutions.face_mesh

# For static images:

import glob
IMAGE_FILES = glob.glob('img.jpg')


drawing_spec = mp_drawing.DrawingSpec(thickness=1, circle_radius=1)
with mp_face_mesh.FaceMesh(
    static_image_mode=True,
    max_num_faces=1,
    refine_landmarks=True,
    min_detection_confidence=0.5) as face_mesh:
  for idx, file in enumerate(IMAGE_FILES):
    image = cv2.imread(file)
    # Convert the BGR image to RGB before processing.
    results = face_mesh.process(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))

    # Print and draw face mesh landmarks on the image.
    if not results.multi_face_landmarks:
      continue
    annotated_image = image.copy()

    landmarks_list = [
        33, 7, 163, 144, 145, 153, 154, 155,
        133, 173, 157, 158, 159, 160, 161, 246
    ]

    for face_landmarks in results.multi_face_landmarks:
      #print('face_landmarks:', face_landmarks)

      print(len(face_landmarks.landmark))
    #   print(face_landmarks.landmark[7])

      for idx in landmarks_list:
        loc_x = int(face_landmarks.landmark[idx].x * image.shape[1])
        loc_y = int(face_landmarks.landmark[idx].y * image.shape[0])
        print(loc_x, loc_y)
        cv2.circle(annotated_image,(loc_x, loc_y), 2, (255,255,255), 2)

    cv2_imshow(annotated_image)

输出

演示笔记本

https://github.com/quickgrid/AI-Resources/blob/master/code-lab/computer_vision/Mediapipe_Python_Holistic_and_Face_Mesh.ipynb