def main():
filterInstance = filter.SimpleMedianFilter()
img = mytools.read("moon.tif")
cv2filter = filterInstance.apply(img)
myapply = filterInstance.myapply(img)
diff = numpy.abs(myapply - cv2filter)
mytools.drawlocalImage([img, cv2filter, myapply, diff], locals())
pass
print("row" + str(counters + 1) + "of" + str(image.shape[0]) + "...")
horizonDFT[counters] = singleRowDft(row)
counters += 1
counters = 0
for col in horizonDFT.T:
print("\t" + "row" + str(counters + 1) + "of" + str(image.shape[0]) +
"...")
verticalAndHorizonDft.T[counters] = singleRowDft(col)
counters += 1
return verticalAndHorizonDft
if __name__ == "__main__":
srcImg = tool.read("car.jpg")
srcImg = cv2.resize(srcImg, (100, 100))
# my频率化图像
mydft = myDFT(srcImg)
#格式化显示
mydft_Draw = numpy.log(numpy.abs(numpy.fft.fftshift(mydft)))
#numpy 频率化图像
npDft = numpy.fft.fft2(srcImg)
npdft_Draw = numpy.log(numpy.abs(numpy.fft.fftshift(npDft)))
diff = (npdft_Draw - mydft_Draw)
plt.subplot(231), plt.imshow(srcImg, 'gray'),
for i in range(rows):
for j in range(cols):
dis = math.sqrt((i - rows / 2)**2 + (j - cols / 2)**2)
if (dis <= d0 or dis >= d1): h = 0
else:
h = 1
dstfftl[i, j] = h * sfftl[i, j]
maskwidth = 10
crow, ccol = int(rows / 2), int(cols / 2)
mask = numpy.ones((rows, cols), numpy.uint8)
mask[0:rows, ccol - maskwidth:ccol + maskwidth] = 0
mask[crow - maskwidth:crow + maskwidth, 0:cols] = 0
cv2.imshow("sddd", mask * 244)
return dstfftl * mask
# 带通滤波 高频和低频不通过 中频通过
if __name__ == "__main__":
filterInstance = BandPassFilter()
img = mytools.read("moon.tif")
img = cv2.resize(img, (400, 400))
# 三个图像分别是 归一化的 源图像傅里叶变换， 叠加滤波器的傅里叶变化， 和目标图像
sfftl, dstsfftl, dst = filterInstance.apply(img)
mytools.drawlocalImage([img, sfftl, dstsfftl, dst, img - dst], locals())
pass
def histogramEqualize(image, maxIntensity):
rows, cols = image.shape
# get cdf from image
cdf, binCenters = exposure.cumulative_distribution(image, maxIntensity)
binCenters = binCenters.tolist()
for i in range(rows):
for j in range(cols):
try:
probability = cdf[binCenters.index(image[i][j])]
except:
probability = 1
image[i][j] = int(probability * maxIntensity)
return image, binCenters, cdf
if __name__ == "__main__":
image = tool.read("sportscar.jpg")
cv2.imshow("image", image)
maxIntensity = 255 # 8 bits
trans, binCenters, cdf = histogramEqualize(image, maxIntensity)
cv2.imshow("trans", trans)
cv2.waitKey(0)