最近在网上看到一个用python实现的200行的换脸程序,觉得挺有意思的,就在自己的电脑上跑了一下,觉得还是挺有趣的,于是自己学习了代码,原代码是实现双眼鼻子和嘴巴区域的转换,我做了一些小改变,实现整个人脸区域的转换。下面就学习一下如何顺利的将此份代码跑起来并同时对代码中的一些函数做一个简单的解释。

1.   环境的配置

笔者是在windows7系统下搭建环境的,采用的是anaconda4.2.0,里面的python版本为python2.7.12。anaconda的安装比较简单,安装完成后运行cmd输入python显示如下图则说明python安装成功。

python deepface 换脸 python opencv换脸_python

下面就是opencv和dlib的安装:

opencv的安装:笔者使用的是opencv2.4.9,点击下载得到的exe文件将opencv解压到某一目录下,比如D:\Program Files\opencv,该目录下有两个文件夹,一个build一个sources,然后将D:\Program Files\opencv\build\python\2.7\x64目录下的cv2.pyd复制到D:\Program Files\anaconda\Lib\site-packages下即可。同样在终端输入python之后import cv2,如果不报错,则说明安装成功。

python deepface 换脸 python opencv换脸_python_02

dlib的安装:最简单的安装方式是敲一条命令即可,打开cmd窗口,输入conda install -c menpo dlib=18.18,如果提示成功安装很好,不然的话估计会很麻烦,笔者没有试过,这样安装完成后同样import dlib,不提示出错则说明安装成功。

python deepface 换脸 python opencv换脸_换脸_03

至此,环境已搭建好,下面就可以使用pycharm运行程序了。

2. 代码的运行与解读

我们需要两个参数即两张图片作为命令行参数,在pycharm中点击Run-Edit Configurations,然后在脚本参数上添加两张图片名,如下图所示

python deepface 换脸 python opencv换脸_dlib_04

然后点击运行即可。


我们先来看一下效果图,下图即为1.jpg为需要换脸的图片(谢谢井宝帅气的照片)

python deepface 换脸 python opencv换脸_dlib_05


下图即2.jpg,表示需要换的人脸

python deepface 换脸 python opencv换脸_python_06


程序运行完生成的图片如下图所示,可以发现人脸区域已经被换掉

python deepface 换脸 python opencv换脸_dlib_07

在这里请求井宝的粉丝不要杀了我,只是用一下他的照片,如果伤害了你,请联系笔者删除,你们的井宝还是帅帅的。


整个代码的执行流程大概如下:

使用dlib提取人脸68个关键点-->用普式分析调整脸部-->校正第二张图片的颜色-->进行两张图片的混合

下面附上所有的代码,代码中加了一些自己理解的注释,可能不准,仅供参考


import cv2
import dlib
import numpy
import sys

PREDICTOR_PATH = "./shape_predictor_68_face_landmarks.dat"
SCALE_FACTOR = 1 
FEATHER_AMOUNT = 11
# 代表各个区域的关键点标号
FACE_POINTS = list(range(17, 68))
MOUTH_POINTS = list(range(48, 61))
RIGHT_BROW_POINTS = list(range(17, 22))
LEFT_BROW_POINTS = list(range(22, 27))
RIGHT_EYE_POINTS = list(range(36, 42))
LEFT_EYE_POINTS = list(range(42, 48))
NOSE_POINTS = list(range(27, 35))
JAW_POINTS = list(range(0, 17))

# Points used to line up the images.   17-61
ALIGN_POINTS = (LEFT_BROW_POINTS + RIGHT_EYE_POINTS + LEFT_EYE_POINTS +
                               RIGHT_BROW_POINTS + NOSE_POINTS + MOUTH_POINTS)

# Points from the second image to overlay on the first. The convex hull of each
# element will be overlaid.   17-61
OVERLAY_POINTS = [
    LEFT_EYE_POINTS + RIGHT_EYE_POINTS + LEFT_BROW_POINTS + RIGHT_BROW_POINTS,
    NOSE_POINTS+ MOUTH_POINTS,
  ]
# Amount of blur to use during colour correction, as a fraction of the
# pupillary distance.
COLOUR_CORRECT_BLUR_FRAC = 0.6

detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(PREDICTOR_PATH)

class TooManyFaces(Exception):
    pass
class NoFaces(Exception):
    pass

# 获取关键点坐标位置,只获取一张人脸
# input:代表一张图片的numpy array
# output:68*2的关键点坐标位置matrix
def get_landmarks(im):
    rects = detector(im, 1)
    if len(rects) > 1:
        raise TooManyFaces
    if len(rects) == 0:
        raise NoFaces
    return numpy.matrix([[p.x, p.y] for p in predictor(im, rects[0]).parts()])

def read_im_and_landmarks(fname):
    im = cv2.imread(fname, cv2.IMREAD_COLOR)
    im = cv2.resize(im, (im.shape[1] * SCALE_FACTOR, im.shape[0] * SCALE_FACTOR))
    s = get_landmarks(im)
    return im, s

# 注解关键点
def annotate_landmarks(im, landmarks):
    # 数组切片是原始数组的视图,这意味着数据不会被复制,视图上的任何修改都会被直接反映到源数组上.
    # 若想要得到的是ndarray切片的一份副本而非视图,就需要显式的进行复制操作函数copy()。
    im = im.copy()
    for idx, point in enumerate(landmarks):
        pos = (point[0, 0], point[0, 1])
        cv2.putText(im, str(idx), pos,
                    fontFace=cv2.FONT_HERSHEY_SCRIPT_SIMPLEX,
                    fontScale=0.2,
                    color=(0, 0, 255))
        cv2.circle(im, pos, 1, color=(0, 255, 255))
        cv2.imwrite("landmak.jpg",im)
    return im

def draw_convex_hull(im, points, color):
    points = cv2.convexHull(points)  # 检测凸包函数
    cv2.fillConvexPoly(im, points, color=color)  # 绘制好多边形后并填充     点的顺序不同绘制出来的凸包也不同

def get_face_mask(im, landmarks):
    im = numpy.zeros(im.shape[:2], dtype=numpy.float64)

    # for group in OVERLAY_POINTS:
    #     draw_convex_hull(im,landmarks[group],color=1)

    # 11. 下面这行代码用来替代上面两行代码
    draw_convex_hull(im,landmarks,color=1)
    im = numpy.array([im, im, im]).transpose((1, 2, 0))  # 得到一个类似于3通道的图片

    # 22. 高斯滤波,注释掉效果更好
    #im = (cv2.GaussianBlur(im, (FEATHER_AMOUNT, FEATHER_AMOUNT), 0) > 0) * 1.0
    #im = cv2.GaussianBlur(im, (FEATHER_AMOUNT, FEATHER_AMOUNT), 0)
    return im

# 用普氏分析(Procrustes analysis)调整脸部
def transformation_from_points(points1, points2):
    """
    Return an affine transformation [s * R | T] such that:返回一个仿射变换矩阵

        sum ||s*R*p1,i + T - p2,i||^2

    is minimized.

    """
    # 通过减去中心id,通过标准偏差进行缩放,然后使用SVD来计算旋转,从而解决了普是问题
    # Solve the procrustes problem by subtracting centroids, scaling by the
    # standard deviation, and then using the SVD to calculate the rotation. See
    # the following for more details:
    #   https://en.wikipedia.org/wiki/Orthogonal_Procrustes_problem

    points1 = points1.astype(numpy.float64)
    points2 = points2.astype(numpy.float64)
    c1 = numpy.mean(points1, axis=0)
    c2 = numpy.mean(points2, axis=0)
    points1 -= c1
    points2 -= c2
    # 计算标准差
    s1 = numpy.std(points1)
    s2 = numpy.std(points2)
    points1 /= s1
    points2 /= s2
    # 通过奇异值分解求得旋转矩阵R
    U, S, Vt = numpy.linalg.svd(points1.T * points2)

    # The R we seek is in fact the transpose of the one given by U * Vt. This
    # is because the above formulation assumes the matrix goes on the right
    # (with row vectors) where as our solution requires the matrix to be on the
    # left (with column vectors).
    R = (U * Vt).T   # 维度:2*2
    # 仿射变换矩阵3*3 #  numpy.hstack用来在第1个维度上拼接tup  numpy.vstack在第0个维度上拼接tup
    return numpy.vstack([numpy.hstack(((s2 / s1) * R,
                                       c2.T - (s2 / s1) * R * c1.T)),
                         numpy.matrix([0., 0., 1.])])

def warp_im(im, M, dshape):
    output_im = numpy.zeros(dshape, dtype=im.dtype)
    # cv2.warpAffine(src, M, dsize[, dst[, flags[, borderMode[, borderValue ]]]])-->dst
    cv2.warpAffine(im,M[:2],(dshape[1], dshape[0]),dst=output_im,borderMode=cv2.BORDER_TRANSPARENT,flags=cv2.WARP_INVERSE_MAP)
    return output_im

# 颜色校正
def correct_colours(im1, im2, landmarks1):
    blur_amount = COLOUR_CORRECT_BLUR_FRAC * numpy.linalg.norm(numpy.mean(landmarks1[LEFT_EYE_POINTS], axis=0)-numpy.mean(landmarks1[RIGHT_EYE_POINTS], axis=0))
    blur_amount = int(blur_amount)
    if blur_amount % 2 == 0:
        blur_amount += 1
    im1_blur = cv2.GaussianBlur(im1, (blur_amount, blur_amount), 0)
    im2_blur = cv2.GaussianBlur(im2, (blur_amount, blur_amount), 0)
    # Avoid divide-by-zero errors.
    im2_blur += (128 * (im2_blur <= 1.0)).astype(im2_blur.dtype)
    return (im2.astype(numpy.float64) * im1_blur.astype(numpy.float64)/im2_blur.astype(numpy.float64))

im1, landmarks1 = read_im_and_landmarks(sys.argv[1])
im2, landmarks2 = read_im_and_landmarks(sys.argv[2])
# 44. 参数landmarks1[ALIGN_POINTS]-->landmarks1
M = transformation_from_points(landmarks1,landmarks2)  # [ALIGN_POINTS]

# get_face_mask()的定义是为一张图像和一个标记矩阵生成一个掩膜
mask = get_face_mask(im2, landmarks2)
warped_mask = warp_im(mask, M, im1.shape)
# 33. 用min函数取掩膜区域效果更好
combined_mask = numpy.min([get_face_mask(im1, landmarks1), warped_mask],axis=0) 
# 将图像2的掩膜转换到图像1的坐标空间
warped_im2 = warp_im(im2, M, im1.shape)
warped_corrected_im2 = correct_colours(im1, warped_im2, landmarks1)
output_im = im1 * (1.0 - combined_mask) + warped_corrected_im2 * combined_mask
cv2.imwrite('output.jpg', output_im)



相比原来的代码,以上代码做了4处改变:

(1)在调用函数transformation_from_points生成仿射矩阵时,我们使用68个关键点作为参数,而不是使用原来的45个关键点

(2)在计算图像掩膜时,去掉高斯滤波操作

(3)在计算combined_mask时,用min函数取代max函数

(4)在计算掩膜时,直接将所有landmark作为参数传进去调用draw_convex_hull(im,landmarks,color=1)