本项目针对疫情期间毕业生无法拍摄毕业照的遗憾，提供DIY毕业照解决方案。通过AI换lian将个人人脸合成到样本图，再经毕业服装抠图与合成、人体抠图与学校背景合成，完成毕业照制作。使用paddlehub等工具实现，但存在服装学科颜色、帽子垂穗处理等需完善的瑕疵，最终祝福毕业生前程似锦。
☞☞☞AI 智能聊天, 问答助手, AI 智能搜索, 免费无限量使用 DeepSeek R1 模型☜☜☜
项目简介
由于疫情的影响，许多毕业生他们可能都没办法拥有一张属于自己的毕业照，这将成为许多人说遗憾。于是便做了这样一个DIY毕业照的项目，最后祝福各位毕业生前程似锦，万事如意。
一、安装必要的包
In [1]
!pip install --upgrade paddlehub -i https://mirror.baidu.com/pypi/simple !hub install deeplabv3p_xception65_humanseg==1.1.2
二、AI换lian
把自己的脸合成到样本图上
只需修改im1，im2
im1：自己脸的图片
im2：样本图
In [ ]
import cv2import numpy as npimport paddlehub as hub
In [ ]
def get_image_size(image):
"""
获取图片大小（高度,宽度）
:param image: image
:return: （高度,宽度）
"""
image_size = (image.shape[0], image.shape[1]) return image_sizedef get_face_landmarks(image):
"""
获取人脸标志，68个特征点
:param image: image
:param face_detector: dlib.get_frontal_face_detector
:param shape_predictor: dlib.shape_predictor
:return: np.array([[],[]]), 68个特征点
"""
dets = face_landmark.keypoint_detection([image])
num_faces = len(dets[0]['data'][0]) if num_faces == 0: print("Sorry, there were no faces found.") return None
# shape = shape_predictor(image, dets[0])
face_landmarks = np.array([[p[0], p[1]] for p in dets[0]['data'][0]]) return face_landmarksdef get_face_mask(image_size, face_landmarks):
"""
获取人脸掩模
:param image_size: 图片大小
:param face_landmarks: 68个特征点
:return: image_mask, 掩模图片
"""
mask = np.zeros(image_size, dtype=np.int32)
points = np.concatenate([face_landmarks[0:16], face_landmarks[26:17:-1]])
points = np.array(points, dtype=np.int32)
cv2.fillPoly(img=mask, pts=[points], color=255) # mask = np.zeros(image_size, dtype=np.uint8)
# points = cv2.convexHull(face_landmarks) # 凸包
# cv2.fillConvexPoly(mask, points, color=255)
return mask.astype(np.uint8)def get_affine_image(image1, image2, face_landmarks1, face_landmarks2):
"""
获取图片1仿射变换后的图片
:param image1: 图片1, 要进行仿射变换的图片
:param image2: 图片2, 只要用来获取图片大小，生成与之大小相同的仿射变换图片
:param face_landmarks1: 图片1的人脸特征点
:param face_landmarks2: 图片2的人脸特征点
:return: 仿射变换后的图片
"""
three_points_index = [18, 8, 25]
M = cv2.getAffineTransform(face_landmarks1[three_points_index].astype(np.float32),
face_landmarks2[three_points_index].astype(np.float32))
dsize = (image2.shape[1], image2.shape[0])
affine_image = cv2.warpAffine(image1, M, dsize) return affine_image.astype(np.uint8)def get_mask_center_point(image_mask):
"""
获取掩模的中心点坐标
:param image_mask: 掩模图片
:return: 掩模中心
"""
image_mask_index = np.argwhere(image_mask > 0)
miny, minx = np.min(image_mask_index, axis=0)
maxy, maxx = np.max(image_mask_index, axis=0)
center_point = ((maxx + minx) // 2, (maxy + miny) // 2) return center_pointdef get_mask_union(mask1, mask2):
"""
获取两个掩模掩盖部分的并集
:param mask1: mask_image, 掩模1
:param mask2: mask_image, 掩模2
:return: 两个掩模掩盖部分的并集
"""
mask = np.min([mask1, mask2], axis=0) # 掩盖部分并集
mask = ((cv2.blur(mask, (5, 5)) == 255) * 255).astype(np.uint8) # 缩小掩模大小
mask = cv2.blur(mask, (3, 3)).astype(np.uint8) # 模糊掩模
return maskdef skin_color_adjustment(im1, im2, mask=None):
"""
肤色调整
:param im1: 图片1
:param im2: 图片2
:param mask: 人脸 mask. 如果存在，使用人脸部分均值来求肤色变换系数；否则，使用高斯模糊来求肤色变换系数
:return: 根据图片2的颜色调整的图片1
"""
if mask is None:
im1_ksize = 55
im2_ksize = 55
im1_factor = cv2.GaussianBlur(im1, (im1_ksize, im1_ksize), 0).astype(np.float)
im2_factor = cv2.GaussianBlur(im2, (im2_ksize, im2_ksize), 0).astype(np.float) else:
im1_face_image = cv2.bitwise_and(im1, im1, mask=mask)
im2_face_image = cv2.bitwise_and(im2, im2, mask=mask)
im1_factor = np.mean(im1_face_image, axis=(0, 1))
im2_factor = np.mean(im2_face_image, axis=(0, 1))
im1 = np.clip((im1.astype(np.float) * im2_factor / np.clip(im1_factor, 1e-6, None)), 0, 255).astype(np.uint8) return im1def main():
im1 = cv2.imread("face.png") # face_image
im1 = cv2.resize(im1, (600, im1.shape[0] * 600 // im1.shape[1]))
landmarks1 = get_face_landmarks(im1) # 68_face_landmarks
if landmarks1 is None: print('{}:检测不到人脸'.format(image_face_path))
exit(1)
im1_size = get_image_size(im1) # 脸图大小
im1_mask = get_face_mask(im1_size, landmarks1) # 脸图人脸掩模
# ret_val, im2 = cam.read() # camera_image
im2 = cv2.imread("di_zhao.png")
landmarks2 = get_face_landmarks(im2) # 68_face_landmarks
if landmarks2 is not None:
im2_size = get_image_size(im2) # 摄像头图片大小
im2_mask = get_face_mask(im2_size, landmarks2) # 摄像头图片人脸掩模
affine_im1 = get_affine_image(im1, im2, landmarks1, landmarks2) # im1（脸图）仿射变换后的图片
affine_im1_mask = get_affine_image(im1_mask, im2, landmarks1, landmarks2) # im1（脸图）仿射变换后的图片的人脸掩模
union_mask = get_mask_union(im2_mask, affine_im1_mask) # 掩模合并
# affine_im1_face_image = cv2.bitwise_and(affine_im1, affine_im1, mask=union_mask) # im1（脸图）的脸
# im2_face_image = cv2.bitwise_and(im2, im2, mask=union_mask) # im2（摄像头图片）的脸
# cv2.imshow('affine_im1_face_image', affine_im1_face_image)
# cv2.imshow('im2_face_image', im2_face_image)
affine_im1 = skin_color_adjustment(affine_im1, im2, mask=union_mask) # 肤色调整
point = get_mask_center_point(affine_im1_mask) # im1（脸图）仿射变换后的图片的人脸掩模的中心点
seamless_im = cv2.seamlessClone(affine_im1, im2, mask=union_mask, p=point, flags=cv2.NORMAL_CLONE) # 进行泊松融合
# cv2.imshow('affine_im1', affine_im1)
# cv2.imshow('im2', im2)
# cv2.imshow('seamless_im', seamless_im)
cv2.imwrite('hecheng.jpg', seamless_im) # plt.imshow(seamless_im)
# plt.show()
else:
cv2.imshow('seamless_im', im2) # plt.imshow(im2)
# plt.show()if __name__ == '__main__':
face_landmark = hub.Module(name="face_landmark_localization")
main()
[2022-06-07 11:22:24,086] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object [2022-06-07 11:22:24,177] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
三、毕业服装抠图与合成
In [ ]
#调用一些相关的包import matplotlibimport matplotlib.pyplot as plt import matplotlib.image as mpimg import cv2from PIL import Imageimport numpy as npimport paddlehub as hub
In [ ]
# S1 衣服图片抠图 ---------------------------------------------------------------------module = hub.Module(name="deeplabv3p_xception65_humanseg")
res = module.segmentation(paths = ["bo.png"], visualization=True, output_dir='pic_output')
res_img_path = './pic_output/bo.png'img = mpimg.imread(res_img_path)
plt.figure(figsize=(10, 10))
plt.imshow(img)
plt.axis('off')
plt.show()
[2022-06-07 16:11:24,181] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
In [ ]
# S2 显示原始图片 ---------------------------------------------------------------------# 原始图片test_img_path = ["hecheng.jpg"]#import numpy as np #wpb addimg = mpimg.imread(test_img_path[0])
# 展示 原始图片plt.figure(figsize=(10,10))
plt.imshow(img) #wpb comment#plt.imshow(img.astype(np.uint8))#wpb addplt.axis('off')
plt.show()
In [ ]
# S3 获取关键点图像 ---------------------------------------------------------------------module = hub.Module(name="human_pose_estimation_resnet50_mpii")
res = module.keypoint_detection(paths = ["hecheng.jpg"], visualization=True, output_dir='pic_output')
res_img_path = './pic_output/hecheng.jpg'img = mpimg.imread(res_img_path)
plt.figure(figsize=(10, 10))
plt.imshow(img)
plt.axis('off')
plt.show()print(res)
[2022-06-07 16:12:03,050] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
image saved in pic_output/hechengtime=1654589524.jpg
[{'path': 'hecheng.jpg', 'data': OrderedDict([('left_ankle', [205, 698]), ('left_knee', [200, 698]), ('left_hip', [211, 490]), ('right_hip', [269, 482]), ('right_knee', [274, 705]), ('right_ankle', [264, 297]), ('pelvis', [227, 490]), ('thorax', [242, 319]), ('upper_neck', [242, 267]), ('head_top', [242, 133]), ('right_wrist', [190, 467]), ('right_elbow', [110, 423]), ('right_shoulder', [153, 319]), ('left_shoulder', [332, 319]), ('left_elbow', [369, 430]), ('left_wrist', [279, 467])])}]
In [ ]
# S4 换衣服 ---------------------------------------------------------------------#获取衣服位置left_posx=res[0]["data"]["right_shoulder"][0]
left_posy=res[0]["data"]["right_shoulder"][1]
right_posx=res[0]["data"]["left_ankle"][0]
right_posy=res[0]["data"]["left_ankle"][1]print(left_posx, left_posy)print(right_posx, right_posy)#读取图片Image1 = Image.open('hecheng.jpg')
Image1copy = Image1.copy()
Image2 = Image.open('pic_output/bo.png')
Image2copy = Image2.copy()
#resize clothes 可以对抠出的服装图片进行放大缩小width,height=Image1copy.size
newsize=(int(width*1.0),int(height*0.9))
Image2copy = Image2.resize(newsize)#制定要粘贴左上角坐标 可以抠出的服装图片进行移动position=(int(left_posx*-0.07),int(left_posy*0.55) ) # ,right_posx, right_posyprint(position)# 换衣服 ， 应该还有更好的方法进行照片合成Image1copy.paste(Image2copy,position,Image2copy) # 将翻转后图像region 粘贴到 原图im 中的 box位置
# 存为新文件 #Image1copy.save('./pic_output/newclothes.png') Image1copy.save('./pic_output/newclothes_bo.jpg')
# 显示穿着新衣的照片img = mpimg.imread('./pic_output/newclothes_bo.jpg')
plt.figure(figsize=(10,10))
plt.imshow(img)
plt.axis('off')
plt.show()
153 319 205 698 (-10, 175)
四、合成学校背景图片
In [2]
import paddlehub as hubimport matplotlib.pyplot as plt import matplotlib.image as mpimg import cv2from PIL import Imageimport numpy as npimport math
In [7]
import paddlehub as hubimport numpy as npimport matplotlib.pyplot as plt
import matplotlib.image as mpimg
#加载预训练模型"deeplabv3p_xception65_humanseghumanseg = hub.Module(name="deeplabv3p_xception65_humanseg")#可以添加多张图片img_path = ["hecheng.jpg"]
results = humanseg.segmentation(data={"image":img_path},visualization=True, output_dir='humanseg_output')#遍历图片抠图结果for i in range(len(img_path)): #呈现原图
img1 = mpimg.imread(img_path[i])
plt.figure(figsize=(10,10))
plt.imshow(img1)
plt.axis('off')
plt.show()
result=results[i] print(result) #打印 抠图结果的数字列表
# print(result["data"].shape)
#以图形方式呈现结果
prediction = result["data"]
plt.imshow(prediction)
plt.show() #运用线性代数实现：使用抠图数据剪切原图
newimg = np.zeros(img1.shape)
newimg[:,:,0] = img1[:,:,0] * (prediction>0)
newimg[:,:,1] = img1[:,:,1] * (prediction>0)
newimg[:,:,2] = img1[:,:,2] * (prediction>0)
newimg = newimg.astype(np.uint8)
# 抠图结果展示
plt.figure(figsize=(10,10))
plt.imshow(newimg)
plt.axis('off')
plt.show()
[2022-06-07 16:48:59,533] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
{'save_path': 'humanseg_output/hecheng.png', 'data': array([[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
...,
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.]], dtype=float32)}
In [8]
base_image = Image.open(f'xuexiao.jpeg').convert('RGB')
fore_image = Image.open(f'humanseg_output/hecheng.png').resize(base_image.size)# 图片加权合成scope_map = np.array(fore_image)[:,:,-1] / 255scope_map = scope_map[:,:,np.newaxis]
scope_map = np.repeat(scope_map, repeats=3, axis=2)
res_image = np.multiply(scope_map, np.array(fore_image)[:,:,:3]) + np.multiply((1-scope_map), np.array(base_image))#保存图片res_image = Image.fromarray(np.uint8(res_image))
res_image.save(f"humanseg_output/hecheng_xue.jpg")print('照片合成完毕')
plt.figure(figsize=(10,10))
plt.imshow(res_image)
plt.axis('off')
plt.show()
照片合成完毕
总结
本次项目主要使用了脸部抠图+合成、衣服抠图+合成、人体抠图+背景合成，这三大块的功能来完成。但是仍然存在瑕疵。例如服装上学科代表的颜色和帽子垂穗颜色不能更换、以及最后合成后的帽子垂穗会消失，这都是后面需要完善的地方。
