def new_binaryMask(frame, x0, y0, width, height):
cv2.rectangle(frame, (x0, y0), (x0 + width, y0 + height), (0, 255, 0)) # 画出截取的手势框图
roi = frame[y0:y0 + height, x0:x0 + width] # 获取手势框图
#cv2.imshow("roi", roi) # 显示手势框图
res = skinMask2(roi) # 进行肤色检测
# cv2.imshow("肤色检测后的图像", res) # 显示肤色检测后的图像
###############服饰膨胀##############################
kernel = np.ones((3, 3), np.uint8) # 设置卷积核
erosion = cv2.erode(res, kernel) # 腐蚀操作
# cv2.imshow("erosion", erosion)
dilation = cv2.dilate(erosion, kernel) # 膨胀操作
# cv2.imshow("dilation", dilation)
#
# ##############轮廓提取#######################################
# ret = new_Ca nny(dilation,res)
# print(dilation.shape)
ret, descirptor_in_use = fd.fourierDesciptor(dilation)
black = fd.reconstruct(dilation,descirptor_in_use)
# print(ret.shape)
# cv2.imshow("black", black)
# cv2.imshow("ret", ret)
# print("fourier_result",fourier_result)
# cv2.imshow("fourier_result", fourier_result)
# print(ret_image.shape)
# cv2.imshow("ret_image", ret_image)
# cv2.imshow("fourier_result", fourier_result)
return ret
def new_binaryMask(frame, x0, y0, width, height):
cv2.rectangle(frame, (x0, y0), (x0 + width, y0 + height), (0, 255, 0)) # 画出截取的手势框图
roi = frame[y0:y0 + height, x0:x0 + width] # 获取手势框图
#cv2.imshow("roi", roi) # 显示手势框图
res = skinMask2(roi) # 进行肤色检测
cv2.imshow("res", res) # 显示肤色检测后的图像
###############服饰膨胀##############################
kernel = np.ones((3, 3), np.uint8) # 设置卷积核
erosion = cv2.erode(res, kernel) # 腐蚀操作
# cv2.imshow("erosion", erosion)
dilation = cv2.dilate(erosion, kernel) # 膨胀操作
# cv2.imshow("dilation", dilation)
##############轮廓提取#######################################
# binaryimg = cv2.Canny(res, 50, 200) # 二值化，canny检测
# h = cv2.findContours(binaryimg, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE) # 寻找轮廓
# contours = h[1] # 提取轮廓
# ret = np.ones(res.shape, np.uint8) # 创建黑色幕布
# cv2.drawContours(ret, contours, -1, (255, 255, 255), 1) # 绘制白色轮廓
# cv2.imshow("ret", ret)
ret, fourier_result = fd.fourierDesciptor(dilation)
#cv2.imshow("ret", ret)
return ret
def binaryMask(frame, x0, y0, width, height):
cv2.rectangle(frame, (x0, y0), (x0 + width, y0 + height), (0, 255, 0))
roi = frame[y0:y0 + height, x0:x0 + width]
res = skinMask(roi) #肤色检测
ret, fourier_result = fd.fourierDesciptor(res)
efd_result, K, T = efd.elliptic_fourier_descriptors(res)
return roi, res, ret, fourier_result, efd_result
def binaryMask(frame, x0, y0, width, height): cv2.rectangle(frame,(x0,y0),(x0+width, y0+height),(0,255,0)) #画出截取的手势框图 roi = frame[y0:y0+height, x0:x0+width] #获取手势框图 res = skinMask(roi) #进行肤色检测 kernel = np.ones((3,3), np.uint8) #设置卷积核 res = cv2.erode(res, kernel) #腐蚀操作 res = cv2.dilate(res, kernel)#膨胀操作 ret, fourier_result = fd.fourierDesciptor(res) return roi, res, ret, fourier_result
def deom1():
cp = cv2.VideoCapture(file)
if not cp.isOpened():
print("video open error!")
fps = cp.get(cv2.CAP_PROP_FPS)
frames = cp.get(cv2.CAP_PROP_FRAME_COUNT)
width = int(cp.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cp.get(cv2.CAP_PROP_FRAME_HEIGHT))
video_writer = cv2.VideoWriter(filename=out_file,
fourcc=fourcc,
fps=fps,
frameSize=(300, 300))
clf = joblib.load(model_path + "svm_efd_" + "train_model.m")
# print(frames)
for i in range(int(frames)):
# for i in range(100):
ret, frame = cp.read()
if not ret:
print("file end")
reframe = frame_resize(frame, width, height)
# video_writer.write(reframe)
# cv2.imwrite('C:/Users/50493/Desktop/reframe.png', reframe)
roi, res = pic.binaryMask(reframe, 0, 0, 300, 300)
# cv2.imwrite('C:/Users/50493/Desktop/reframe.png', res)
img, d = fd.fourierDesciptor(res)
descirptor_in_use = abs(d)
temp = descirptor_in_use[1]
feature = []
for k in range(1, len(descirptor_in_use)):
x_record = int(100 * descirptor_in_use[k] / temp)
feature.append(x_record)
returnVec = np.zeros((1, N))
for i in range(N):
returnVec[0, i] = int(feature[i])
valTest = clf.predict(returnVec)
if valTest == 11:
cv2.putText(roi, "like", (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1.0,
(0, 0, 255), 1)
video_writer.write(roi)
def binaryMask(frame, x0, y0, width, height):
cv2.rectangle(frame, (x0, y0), (x0 + width, y0 + height),
(0, 255, 0)) #画出截取的手势框图
roi = frame[y0:y0 + height, x0:x0 + width] #获取手势框图
#roi = topolar(roi)
#cv2.imshow("roi", roi) #显示手势框图
res = skinMask(roi) #进行肤色检测
#cv2.imshow("res", res) #显示肤色检测后的图像
'''
kernel = np.ones((3,3), np.uint8) #设置卷积核
res = cv2.erode(res, kernel) #腐蚀操作
res = cv2.dilate(res, kernel)#膨胀操作
cv2.imshow("res", res) #显示肤色检测后的图像
'''
ret, fourier_result = fd.fourierDesciptor(res)
efd_result, K, T = efd.elliptic_fourier_descriptors(res)
#ret = fd.reconstruct(res, fourier_result, MIN_DESCRIPTOR)
return roi, res, ret, fourier_result, efd_result
def deom2():
cameraCapture = cv2.VideoCapture(0)
cameraCapture.open(0)
cv2.namedWindow('show')
fps = cameraCapture.get(cv2.CAP_PROP_FPS)
frames = cameraCapture.get(cv2.CAP_PROP_FRAME_COUNT)
width = int(cameraCapture.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cameraCapture.get(cv2.CAP_PROP_FRAME_HEIGHT))
# video_writer = cv2.VideoWriter(filename=out_file, fourcc=fourcc, fps=fps,
# frameSize=(300, 300))
clf = joblib.load(model_path + "svm_efd_" + "train_model.m")
ret, frame = cameraCapture.read()
while ret:
cv2.waitKey(1)
ret, frame = cameraCapture.read()
reframe = frame_resize(frame, width, height)
roi, res = pic.binaryMask(reframe, 0, 0, 300, 300)
img, d = fd.fourierDesciptor(res)
descirptor_in_use = abs(d)
temp = descirptor_in_use[1]
feature = []
for k in range(1, len(descirptor_in_use)):
x_record = int(100 * descirptor_in_use[k] / temp)
feature.append(x_record)
returnVec = np.zeros((1, N))
for i in range(N):
returnVec[0, i] = int(feature[i])
valTest = clf.predict(returnVec)
if valTest == 11:
cv2.putText(roi, "like", (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1.0,
(0, 0, 255), 1)
# video_writer.write(roi)
cv2.imshow('show', roi)
cameraCapture.release()
#!/usr/bin/env python
# -*-coding:utf-8 -*-
import efd
import fourierDescriptor as fd
import cv2
import numpy as np
path = './' + 'feature' + '/'
path_img = './' + 'image' + '/'
if __name__ == "__main__":
for i in range(1, 11):
for j in range(1, 201):
roi = cv2.imread(path_img + str(i) + '_' + str(j) + '.png')
descirptor_in_use = abs(fd.fourierDesciptor(roi))
fd_name = path + str(i) + '_' + str(j) + '.txt'
# fd_name = path + str(i) + '.txt'
with open(fd_name, 'w', encoding='utf-8') as f:
temp = descirptor_in_use[1]
for k in range(1, len(descirptor_in_use)):
x_record = int(100 * descirptor_in_use[k] / temp)
f.write(str(x_record))
f.write(' ')
f.write('\n')
print(i, '_', j, '完成')
'''
for i in range(1, 11):
for j in range(1, 21):
roi = cv2.imread(path_img + str(i) + '_' + str(j) + '.png')