-
Notifications
You must be signed in to change notification settings - Fork 0
/
chapter5.py
180 lines (169 loc) · 7.32 KB
/
chapter5.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
# -*- coding:UTF-8 -*-
# !/usr/bin/python
import numpy as np
import cv2
import os
import time
# 静态图像人脸检测
# def detect(img_path):
# face_cascade = cv2.CascadeClassifier("haarcascade/haarcascade_frontalface_default.xml") # 分类器
# img = cv2.imread(img_path)
# gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# faces = face_cascade.detectMultiScale(gray_img, 1.3, 5) # 在灰度图像上人脸检测(多个坐标点)
# for x, y, w, h in faces: # 根据坐标点在彩图上画矩形
# img = cv2.rectangle(img=img, pt1=(x, y), pt2=(x + w, y + h), color=(255, 0, 0), thickness=2)
# cv2.namedWindow("face")
# cv2.imshow("face", img)
# cv2.waitKey(10000)
#
# detect("img/ice&fire2.jpg")
# 动态图像检测（摄像头检测人脸模型）
cascade_file_hash = {
"face": cv2.CascadeClassifier("haarcascade/haarcascade_frontalface_default.xml"),
"l_eye": cv2.CascadeClassifier("haarcascade/haarcascade_mcs_lefteye_alt.xml"),
"r_eye": cv2.CascadeClassifier("haarcascade/haarcascade_mcs_righteye_alt.xml"),
"mouth": cv2.CascadeClassifier("haarcascade/haarcascade_mcs_mouth.xml"),
"nose": cv2.CascadeClassifier("haarcascade/haarcascade_mcs_nose.xml")
}
face_label_hash = {1: "rex", 2: "iris"}
face_label_hash_ = {"rex": 1, "iris": 2}
# 0.人脸检测demo
# def face_detect(face_frame, cascade_file):
# gray = cv2.cvtColor(face_frame, cv2.COLOR_BGR2GRAY)
# # 脸部轮廓
# faces_coord = cascade_file["face"].detectMultiScale(gray, 1.3, 5)
# for x, y, w, h in faces_coord:
# face_frame = cv2.rectangle(face_frame, pt1=(x, y), pt2=(x + w, y + h), color=(0, 255, 0), thickness=2)
# # # 左眼睛
# # leye_coord = cascade_file["l_eye"].detectMultiScale(gray, 1.3, 5)
# # for x, y, w, h in leye_coord:
# # face_frame = cv2.rectangle(face_frame, pt1=(x, y), pt2=(x + w, y + h), color=(0, 255, 255), thickness=2)
# # # 右眼睛
# # reye_coord = cascade_file["r_eye"].detectMultiScale(gray, 1.3, 5)
# # for x, y, w, h in reye_coord:
# # face_frame = cv2.rectangle(face_frame, pt1=(x, y), pt2=(x + w, y + h), color=(0, 255, 255), thickness=2)
# # # 嘴巴
# # reye_coord = cascade_file["mouth"].detectMultiScale(gray, 1.3, 5)
# # for x, y, w, h in reye_coord:
# # face_frame = cv2.rectangle(face_frame, pt1=(x, y), pt2=(x + w, y + h), color=(255, 0, 0), thickness=2)
# # 鼻子
# reye_coord = cascade_file["nose"].detectMultiScale(gray, 1.3, 5)
# for x, y, w, h in reye_coord:
# face_frame = cv2.rectangle(face_frame, pt1=(x, y), pt2=(x + w, y + h), color=(0, 0, 255), thickness=2)
# return face_frame
#
# cc = cv2.VideoCapture(0)
# cv2.namedWindow("face", cv2.WINDOW_AUTOSIZE)
# success, frame = cc.read() # 读取摄像头图像
# while success:
# frame = face_detect(frame, cascade_file_hash) # 检测人脸
# cv2.imshow("face", frame)
# keyCode = cv2.waitKey(1)
# if keyCode == 27:
# break
# success, frame = cc.read() # 继续读取新图像
# cv2.destroyAllWindows()
# cc.release() # 释放资源
def make_dir(path):
if os.path.isdir(path):
for f in os.listdir(path): os.remove(os.path.join(path, f))
else:
os.mkdir(path)
def time_name(path):
name = time.strftime("%Y%m%d_%H%M%S", time.localtime())
return '%s\\%s.jpg' % (path, name)
# 1.生成人脸识别数据
# name, count = 'iris', 0
# path = 'img\\face\\%s' % name
# make_dir(path)
# cc = cv2.VideoCapture(0)
# cv2.namedWindow("face", cv2.WINDOW_AUTOSIZE)
# success, frame = cc.read()
# while success:
# frame = cv2.flip(frame, flipCode=1) # Y轴翻转
# gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
# coord = cascade_file_hash["face"].detectMultiScale(gray, 1.3, 5)
# if len(coord): # 如果有检测结果
# count += 1
# x, y, w, h = coord[0][0], coord[0][1], coord[0][2], coord[0][3] # 第一人脸矩形参数
# frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
# if count % 10 == 0: # 不连续保存
# frame_scale = gray[x:x + w, y:y + h] # 头像部分
# frame_scale = cv2.resize(frame_scale, (200, 200), interpolation=cv2.INTER_NEAREST) # 统一大小
# cv2.imwrite(path + '\\%d.jpg' % (count // 10), frame_scale) # 保存图像
# cv2.imshow("face", frame)
# keyCode = cv2.waitKey(1)
# if keyCode == 27: break
# success, frame = cc.read()
# cv2.destroyAllWindows()
# cc.release()
# 2.训练数据准备
# # 摄像头数据
# dirnames = os.listdir('img\\face')
# with open("face_train1.csv", "w+") as f:
# for dirs in dirnames:
# dirname = 'img\\face\\' + dirs
# if os.path.isdir(dirname):
# files = os.listdir(dirname)
# for file in files:
# filename = '%s\\%s' % (dirname, file)
# if os.path.isfile(filename): f.writelines("%s;%d\n" % (filename, face_label_hash_[dirs]))
# # 人脸图片数据
# lst, count = [], 0
# for dirname, dirnames, filenames in os.walk("img\\att_faces"): # 广度优先遍历原理
# for subdir in dirnames:
# sublst, subpath = [], os.path.join(dirname, subdir)
# for f in os.listdir(subpath):
# sublst.append(os.path.join(subpath, f))
# lst.append(sublst)
# break
# with open("face_train.csv", "w+") as f:
# for sub in lst:
# for i in sub:
# f.writelines(i + ";%d\n" % count)
# count += 1
# 3.加载训练数据
face_mat, face_label = [], []
with open("face_train1.csv", 'r') as f:
for i in f.readlines():
i_ = i.split(';')
if os.path.exists(i_[0]):
i_[1] = i_[1].replace('\n', '')
i_[0] = cv2.imread(i_[0], cv2.IMREAD_GRAYSCALE)
i_[0] = cv2.resize(i_[0], (200, 200))
face_mat.append(i_[0])
face_label.append(int(i_[1]))
face_mat = np.asarray(face_mat, dtype=np.uint8)
face_label = np.asarray(face_label, dtype=np.int32)
# 4.人脸识别训练与预测(threshold为置信度阈值)
# 训练
model = cv2.face.LBPHFaceRecognizer_create(radius=1, neighbors=8, grid_x=8, grid_y=8, threshold=123.0)
model.train(face_mat, face_label)
# 测试
cc = cv2.VideoCapture(0)
cv2.namedWindow("face", cv2.WINDOW_AUTOSIZE)
success, frame = cc.read()
while success:
frame = cv2.flip(frame, flipCode=1)
gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces_coord = cascade_file_hash["face"].detectMultiScale(gray_frame, 1.3, 5)
for x, y, w, h in faces_coord:
frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
frame = cv2.circle(frame, center=(x, y), radius=4, color=(0, 0, 255), thickness=2)
frame = cv2.circle(frame, center=(x + w, y + h), radius=4, color=(0, 0, 255), thickness=2)
roi = gray_frame[x:x + w, y:y + h]
roi = cv2.resize(roi, (200, 200), interpolation=cv2.INTER_LINEAR)
param = model.predict(roi)
if param[0] != -1:
title = '%s: %d' % (face_label_hash[param[0]], param[1])
cv2.putText(img=frame, text=title, org=(x, y - 20),
fontFace=cv2.FONT_HERSHEY_SIMPLEX, fontScale=0.8, color=(0, 0, 255), thickness=2)
cv2.imshow("face", frame)
keyCode = cv2.waitKey(1)
if keyCode == 27:
break
elif keyCode == 13:
cv2.imwrite(time_name('img\\face'), frame)
success, frame = cc.read()
cv2.destroyAllWindows()
cc.release()