转自:github地址:https://github.com/tomoncle/face-detection-induction-course
目录
简介
摄像头实时运行
图片生成gif动图
在github看到的一个搞笑的小程序分享给大家,github地址:https://github.com/tomoncle/face-detection-induction-course
效果如下:
代码如下:
其中的人脸识别数据可以在此下载:https://jaist.dl.sourceforge.net/project/dclib/dlib/v18.10/shape_predictor_68_face_landmarks.dat.bz2
解压后与程序放在同一文件夹下即可。
程序运行后按q退出,按d加墨镜。
#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time : 2019/3/16 10:10
# @Author : tomoncle
# @Site : https://github.com/tomoncle/face-detection-induction-course
# @File : input_video_stream_paste_mask.py
# @Software: PyCharm
import cv2
import numpy as np
from PIL import Image
from imutils import face_utils, resize
from time import sleep
try:
from dlib import get_frontal_face_detector, shape_predictor
except ImportError:
raise
class DynamicStreamMaskService(object):
"""
动态黏贴面具服务
"""
def __init__(self, saved=False):
self.saved = saved # 是否保存图片
self.listener = True # 启动参数
self.video_capture = cv2.VideoCapture(0) # 调用本地摄像头
self.doing = False # 是否进行面部面具
self.speed = 0.4 # 面具移动速度
self.detector = get_frontal_face_detector() # 面部识别器
self.predictor = shape_predictor("shape_predictor_68_face_landmarks.dat") # 面部分析器
self.fps = 4 # 面具存在时间基础时间
self.animation_time = 0 # 动画周期初始值
self.duration = self.fps * 4 # 动画周期最大值
self.fixed_time = 4 # 画图之后,停留时间
self.max_width = 500 # 图像大小
self.deal, self.text, self.cigarette = None, None, None # 面具对象
def read_data(self):
"""
从摄像头获取视频流,并转换为一帧一帧的图像
:return: 返回一帧一帧的图像信息
"""
_, data = self.video_capture.read()
return data
def save_data(self, draw_img):
"""
保存图片到本地
:param draw_img:
:return:
"""
if not self.saved:
return
draw_img.save("images\\%05d.png" % self.animation_time)
def init_mask(self):
"""
加载面具
:return:
"""
self.console("加载面具...")
self.deal, self.text, self.cigarette = (
Image.open(x) for x in ["images\\deals.png", "images\\text.png", "images\\cigarette.png"]
)
def get_glasses_info(self, face_shape, face_width):
"""
获取当前面部的眼镜信息
:param face_shape:
:param face_width:
:return:
"""
left_eye = face_shape[36:42]
right_eye = face_shape[42:48]
left_eye_center = left_eye.mean(axis=0).astype("int")
right_eye_center = right_eye.mean(axis=0).astype("int")
y = left_eye_center[1] - right_eye_center[1]
x = left_eye_center[0] - right_eye_center[0]
eye_angle = np.rad2deg(np.arctan2(y, x))
deal = self.deal.resize(
(face_width, int(face_width * self.deal.size[1] / self.deal.size[0])),
resample=Image.LANCZOS)
deal = deal.rotate(eye_angle, expand=True)
deal = deal.transpose(Image.FLIP_TOP_BOTTOM)
left_eye_x = left_eye[0, 0] - face_width // 4
left_eye_y = left_eye[0, 1] - face_width // 6
return {"image": deal, "pos": (left_eye_x, left_eye_y)}
def get_cigarette_info(self, face_shape, face_width):
"""
获取当前面部的烟卷信息
:param face_shape:
:param face_width:
:return:
"""
mouth = face_shape[49:68]
mouth_center = mouth.mean(axis=0).astype("int")
cigarette = self.cigarette.resize(
(face_width, int(face_width * self.cigarette.size[1] / self.cigarette.size[0])),
resample=Image.LANCZOS)
x = mouth[0, 0] - face_width + int(16 * face_width / self.cigarette.size[0])
y = mouth_center[1]
return {"image": cigarette, "pos": (x, y)}
def orientation(self, rects, img_gray):
"""
人脸定位
:return:
"""
faces = []
for rect in rects:
face = {}
face_shades_width = rect.right() - rect.left()
predictor_shape = self.predictor(img_gray, rect)
face_shape = face_utils.shape_to_np(predictor_shape)
face['cigarette'] = self.get_cigarette_info(face_shape, face_shades_width)
face['glasses'] = self.get_glasses_info(face_shape, face_shades_width)
faces.append(face)
return faces
def start(self):
"""
启动程序
:return:
"""
self.console("程序启动成功.")
self.init_mask()
while self.listener:
frame = self.read_data()
frame = resize(frame, width=self.max_width)
img_gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
rects = self.detector(img_gray, 0)
faces = self.orientation(rects, img_gray)
draw_img = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
if self.doing:
self.drawing(draw_img, faces)
self.animation_time += self.speed
self.save_data(draw_img)
if self.animation_time > self.duration:
self.doing = False
self.animation_time = 0
else:
frame = cv2.cvtColor(np.asarray(draw_img), cv2.COLOR_RGB2BGR)
cv2.imshow("hello mask", frame)
self.listener_keys()
def listener_keys(self):
"""
设置键盘监听事件
:return:
"""
key = cv2.waitKey(1) & 0xFF
if key == ord("q"):
self.listener = False
self.console("程序退出")
sleep(1)
self.exit()
if key == ord("d"):
self.doing = not self.doing
def exit(self):
"""
程序退出
:return:
"""
self.video_capture.release()
cv2.destroyAllWindows()
def drawing(self, draw_img, faces):
"""
画图
:param draw_img:
:param faces:
:return:
"""
for face in faces:
if self.animation_time < self.duration - self.fixed_time:
current_x = int(face["glasses"]["pos"][0])
current_y = int(face["glasses"]["pos"][1] * self.animation_time / (self.duration - self.fixed_time))
draw_img.paste(face["glasses"]["image"], (current_x, current_y), face["glasses"]["image"])
cigarette_x = int(face["cigarette"]["pos"][0])
cigarette_y = int(face["cigarette"]["pos"][1] * self.animation_time / (self.duration - self.fixed_time))
draw_img.paste(face["cigarette"]["image"], (cigarette_x, cigarette_y),
face["cigarette"]["image"])
else:
draw_img.paste(face["glasses"]["image"], face["glasses"]["pos"], face["glasses"]["image"])
draw_img.paste(face["cigarette"]["image"], face["cigarette"]["pos"], face["cigarette"]["image"])
draw_img.paste(self.text, (75, draw_img.height // 2 + 128), self.text)
@classmethod
def console(cls, s):
print("{} !".format(s))
if __name__ == '__main__':
ms = DynamicStreamMaskService()
ms.start()
#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time : 2019/3/13 14:47
# @Author : tomoncle
# @Site : https://github.com/tomoncle/face-detection-induction-course
# @File : input_static_pic_to_gif2.py
# 参考信息:https://www.makeartwithpython.com/blog/deal-with-it-generator-face-recognition/
# 描述:
# 程序从命令行参数获取图片信息,然后,它将使用Dlib中的人脸检测算法来查看是否有人脸存在。
# 如果有,它将为每个人脸创建一个结束位置,眼镜和烟卷会移动到那里结束。
#
# 然后我们需要缩放和旋转我们的眼镜以适合每个人的脸。
# 我们将使用从Dlib的68点模型返回的点集来找到眼睛的中心,并为它们之间的空间旋转。
#
# 在我们找到眼镜的最终位置和旋转后,我们可以为gif制作动画,眼镜从屏幕顶部进入。
# 我们将使用MoviePy和一个自定义的FaceDetect工具类绘制它。
#
# 同理烟卷也是这样。
#
# 应用程序的体系结构非常简单。我们首先接收图片,然后将其转换为灰度NumPy数组。
# 假如没有人脸,程序会自己退出,如果存在,我们就可以将检测到的人脸信息传递到人脸方向预测模型中。
#
# 通过返回的脸部方向,我们可以选择眼睛,缩放和旋转我们的眼镜框架以适合人的面部大小。
#
# 当然这个程序不仅仅只针对于一张人脸,可以检测多个人脸信息。
#
# 最后,通过获取的人脸列表,我们可以使用MoviePy创建一个绘图,然后生成我们的动画gif。
import moviepy.editor as mpy
import numpy as np
from PIL import Image
from imutils import face_utils
try:
from dlib import get_frontal_face_detector, shape_predictor
except ImportError:
raise
class FaceDetect(object):
def __init__(self, img_src, gif_path=None):
self.gif_max_width = 500
self.duration = 4
self.image = self.load(img_src).convert('RGBA')
self.img_gray = None
self.rects = None
self.deal = None
self.text = None
self.cigarette = None
if not self.validate:
print("没有检测到人脸,程序退出.")
exit(1)
self.init_mask()
self.make_gif(gif_path=gif_path)
@property
def validate(self):
"""
验证是否存在人脸,如果不存在返回False
:return:
"""
if self.image.size[0] > self.gif_max_width:
scaled_height = int(self.gif_max_width * self.image.size[1] / self.image.size[0])
self.image.thumbnail((self.gif_max_width, scaled_height))
self.img_gray = np.array(self.image.convert('L'))
self.rects = self.detector(self.img_gray, 0)
return len(self.rects) > 0
@classmethod
def load(cls, img_src):
"""
加载图片转为Image对象
:param img_src:
:return:
"""
return Image.open(img_src)
@property
def detector(self):
"""
检测是否有人脸
:return:
"""
return get_frontal_face_detector()
@property
def predictor(self):
"""
预测我们的面部方向
:return:
"""
return shape_predictor('shape_predictor_68_face_landmarks.dat')
def init_mask(self):
"""
加载面具
:return:
"""
self.deal, self.text, self.cigarette = (
self.load(x) for x in ["images\\deals.png", "images\\text.png", "images\\cigarette.png"]
)
def get_glasses_info(self, face_shape, face_width):
"""
获取当前面部的眼镜信息
:param face_shape:
:param face_width:
:return:
"""
left_eye = face_shape[36:42]
right_eye = face_shape[42:48]
left_eye_center = left_eye.mean(axis=0).astype("int")
right_eye_center = right_eye.mean(axis=0).astype("int")
y = left_eye_center[1] - right_eye_center[1]
x = left_eye_center[0] - right_eye_center[0]
eye_angle = np.rad2deg(np.arctan2(y, x))
deal = self.deal.resize(
(face_width, int(face_width * self.deal.size[1] / self.deal.size[0])),
resample=Image.LANCZOS)
deal = deal.rotate(eye_angle, expand=True)
deal = deal.transpose(Image.FLIP_TOP_BOTTOM)
left_eye_x = left_eye[0, 0] - face_width // 4
left_eye_y = left_eye[0, 1] - face_width // 6
return {"image": deal, "pos": (left_eye_x, left_eye_y)}
def get_cigarette_info(self, face_shape, face_width):
"""
获取当前面部的烟卷信息
:param face_shape:
:param face_width:
:return:
"""
mouth = face_shape[49:68]
mouth_center = mouth.mean(axis=0).astype("int")
cigarette = self.cigarette.resize(
(face_width, int(face_width * self.cigarette.size[1] / self.cigarette.size[0])),
resample=Image.LANCZOS)
x = mouth[0, 0] - face_width + int(16 * face_width / self.cigarette.size[0])
y = mouth_center[1]
return {"image": cigarette, "pos": (x, y)}
def orientation(self):
"""
人脸定位
:return:
"""
faces = []
for rect in self.rects:
face = {}
face_shades_width = rect.right() - rect.left()
predictor_shape = self.predictor(self.img_gray, rect)
face_shape = face_utils.shape_to_np(predictor_shape)
face['cigarette'] = self.get_cigarette_info(face_shape, face_shades_width)
face['glasses'] = self.get_glasses_info(face_shape, face_shades_width)
faces.append(face)
return faces
def drawing(self, t):
"""
动态画图
:param t:
:return:
"""
draw_img = self.image.convert('RGBA')
if t == 0:
return np.asarray(draw_img)
for face in self.orientation():
if t <= self.duration - 2:
current_x = int(face["glasses"]["pos"][0])
current_y = int(face["glasses"]["pos"][1] * t / (self.duration - 2))
draw_img.paste(face["glasses"]["image"], (current_x, current_y), face["glasses"]["image"])
cigarette_x = int(face["cigarette"]["pos"][0])
cigarette_y = int(face["cigarette"]["pos"][1] * t / (self.duration - 2))
draw_img.paste(face["cigarette"]["image"], (cigarette_x, cigarette_y), face["cigarette"]["image"])
else:
draw_img.paste(face["glasses"]["image"], face["glasses"]["pos"], face["glasses"]["image"])
draw_img.paste(face["cigarette"]["image"], face["cigarette"]["pos"], face["cigarette"]["image"])
draw_img.paste(self.text, (75, draw_img.height // 2 + 128), self.text)
return np.asarray(draw_img)
def make_gif(self, gif_path=None):
"""
:param gif_path: 保存路径
:return:
"""
gif_path = gif_path or "deal.gif"
animation = mpy.VideoClip(self.drawing, duration=self.duration)
animation.write_gif(gif_path, fps=self.duration)
if __name__ == '__main__':
FaceDetect('timg.jpg')