def main():
imageDir = r"C:\Study\test\bone\pic-test"
pmSavingDir = r"C:\Study\test\bone\preprocess\pm-results"
claheSavingDir = r"C:\Study\test\bone\preprocess\clahe-results"
cropSavigDir = r"C:\Study\test\bone\preprocess\crop-results"
util.mkdirs([pmSavingDir, claheSavingDir, cropSavigDir])
picFiles = api.getFiles(imageDir)
total = len(picFiles)
for i, f in enumerate(picFiles):
imgFileName = os.path.basename(f)
print("[info] process {} / {}, img: {}".format(i + 1, total,
imgFileName))
img = cv2.imread(f, 0)
# # 去噪
denoiseImg = PmDenoise(img)
denoiseSavingPath = os.path.join(pmSavingDir, imgFileName)
cv2.imwrite(denoiseSavingPath, denoiseImg)
# 增强
denoiseImg = cv2.imread(denoiseSavingPath, cv2.IMREAD_GRAYSCALE)
claheImg = CLAHE(denoiseImg)
cv2.imwrite(os.path.join(claheSavingDir, imgFileName), claheImg)
# 为方便函数调用，最后统一裁剪
crop(claheSavingDir, cropSavigDir)
def handle(dirs, out_dir, clip):
start_time = datetime.datetime.now()
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
files = getFiles(dirs)
total = len(files)
fail, success, skip, count = 0, 0, 0, 0
for f in files:
count += 1
print(count, '/', total)
img_dirs = os.path.join(out_dir, f.split("\\")[-1])
if os.path.isfile(img_dirs):
skip += 1
continue
try:
# 读取图片
img = cv2.imread(f, 0)
# 切边
x, w, y, h = clip
img = img[x:w, y:h]
# 去除噪点
img = moveNoise(img, 7)
# 根据最大熵算法获得最佳阈值
threshed = maxEntrop(img)
# 二值化
threshold, thrshed_img = cv2.threshold(img, threshed * 0.4, 255,
cv2.THRESH_BINARY)
saveImage(img_dirs, "_threshed_raw", thrshed_img)
# 使用区域生长法分割
img_segement, thresh_img = regionGrowing(img, thrshed_img)
saveImage(img_dirs, "_threshed", thresh_img)
# 去除多余边缘
img_remove_margin = moveMargin(img_segement, thresh_img)
saveImage(img_dirs, "_remove_margin", img_remove_margin)
# 扩充为正方形并缩小为256x256
img_new = normalization(img_remove_margin)
saveImage(img_dirs, "_new", img_new)
# 打印信息到输出台
printToConsole(start_time, f, count, total, 5)
success += 1
except Exception as e:
# 错误情况
saveError(e, out_dir, f)
fail += 1
end_time = datetime.datetime.now()
expend = end_time - start_time
print(
"\n\ntotal: %d\nsuccessful: %d\nskip: %d\nfailed: %d\nExpend time: %s"
% (total, success, skip, fail, expend))
os.startfile(out_dir)
def testPM():
imageDir = r"C:\Study\test\bone\pic-test"
outputDir = r"C:\Study\test\bone\pm-results"
util.mkdirs(outputDir)
picFiles = api.getFiles(imageDir)
total = len(picFiles)
for it, file in enumerate(picFiles):
print("[info] process {} / {}.".format(it + 1, total))
img = cv2.imread(file, 0)
retPic = PmDenoise(img, iter=2)
savePath = os.path.join(outputDir, os.path.basename(file))
# print(savePath)
cv2.imwrite(savePath, retPic)
def handle(dirs, out_dir, clip):
start_time = datetime.datetime.now()
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
files = getFiles(dirs)
total = len(files)
fail, success, skip, count = 0, 0, 0, 0
for f in files:
count += 1
print(count, '/', total)
img_dirs = os.path.join(out_dir, f.split("\\")[-1])
if os.path.isfile(img_dirs):
skip += 1
continue
try:
# 读取图片
img = cv2.imread(f, 0)
# 切边
x, w, y, h = clip
img = img[x:w, y:h]
# 去除噪点
img = moveNoise(img, 7)
# 获得图像像素均值
threshed = getMean(img)
# 二值化
threshold, thrshed_img = cv2.threshold(img, threshed * 0.86, 255,
cv2.THRESH_BINARY)
# 使用轮廓法(速度快)分割
img_segement, thresh_img = maxContour(img, thrshed_img)
# 保存
saveImage(img_dirs, "_new", img_segement)
# 打印信息到输出台
printToConsole(start_time, f, count, total, 5)
success += 1
except Exception as e:
# 错误情况
saveError(e, out_dir, f)
fail += 1
end_time = datetime.datetime.now()
expend = end_time - start_time
print(
"\n\ntotal: %d\nsuccessful: %d\nskip: %d\nfailed: %d\nExpend time: %s"
% (total, success, skip, fail, expend))
os.startfile(out_dir)
def handle(dirs, out_dir, clip, w0):
start_time = datetime.datetime.now()
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
files = getFiles(dirs)
total = len(files)
fail, success, skip, count = 0, 0, 0, 0
for f in files:
count += 1
print(count, '/', total)
img_dirs = os.path.join(out_dir, f.split("\\")[-1])
if os.path.isfile(img_dirs):
skip += 1
continue
try:
# 读取图片
img = cv2.imread(f, 0)
# 切边
x, w, y, h = clip
img = img[x:w, y:h]
# 去除噪点
thresh_value = getThreshValuebyHistogram(img, w0, is_handle=False)
# 二值化
threshold, thrshed_img = cv2.threshold(img, thresh_value, 255,
cv2.THRESH_BINARY)
# 使用区域生长法分割
img_segement, thresh_img = regionGrowing(img, thrshed_img)
# 保存
saveImage(img_dirs, "_new", img_segement)
# 打印信息到输出台
printToConsole(start_time, f, count, total, 5)
success += 1
except Exception as e:
# 错误情况
saveError(e, out_dir, f)
fail += 1
end_time = datetime.datetime.now()
expend = end_time - start_time
print(
"\n\ntotal: %d\nsuccessful: %d\nskip: %d\nfailed: %d\nExpend time: %s"
% (total, success, skip, fail, expend))
os.startfile(out_dir)
def cropImgUsingMarginInfo(marginInfoPath, imageDir, outputDir):
"""切边, 需保证待切边图片与marginInfo中的图片名称一致"""
files = api.getFiles(imageDir)
total = len(files)
marginInfo = pickle.load(open(marginInfoPath,'rb'))
util.mkdirs(outputDir)
for i, f in enumerate(files):
basename = os.path.basename(f)
print("process {} / {}: {}".format(i+1, total, basename))
img = cv2.imread(f, 0)
if basename not in marginInfo:
print("process {} / {}: {}, skip cause current pic name does not find in marginInfo".format(i+1, total, basename))
continue
margin = marginInfo[basename]
retImg = crop.cropImage(img, margin)
cv2.imwrite(os.path.join(outputDir, basename), retImg)
def handle(dirs, out_dir):
start_time = datetime.datetime.now()
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
files = getFiles(dirs)
total = len(files)
fail, success, skip, count = 0, 0, 0, 0
for f in files:
count += 1
print(count, '/', total)
img_dirs = os.path.join(out_dir, f.split("\\")[-1])
if os.path.isfile(img_dirs):
skip += 1
continue
try:
img = cv2.imread(f, 0)
# img_new = moveMargin(img, img)
img_new = normalization(img_new)
saveImage(img_dirs, "", img_new)
# 打印信息到输出台
printToConsole(start_time, f, count, total, 5)
success += 1
except Exception as e:
# 错误情况
saveError(e, out_dir, f)
fail += 1
end_time = datetime.datetime.now()
expend = end_time - start_time
print(
"\n\ntotal: %d\nsuccessful: %d\nskip: %d\nfailed: %d\nExpend time: %s"
% (total, success, skip, fail, expend))
os.startfile(out_dir)
def handle(dirs, out_dir, clip):
"""
随机选取20张图片,在像素平均值附近取值
dirs: 原图路径,20张图片
out_dir: 二值图片输出路径
clip = (30,-30,30,-30)
"""
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
files = getFiles(dirs)
# 在label.txt里加上阈值
out_label = os.path.join(out_dir, 'label.txt')
label_file = open(out_label, "w")
out_data = os.path.join(out_dir, 'data.txt')
data_file = open(out_data, "w")
# 原图对应序号
record = os.path.join(out_dir, 'record.txt')
record_file = open(record, "w")
count = 0
for f in files:
record_data = str(count) + '\t' + str(f) + '\n'
record_file.write(record_data)
# continue
img = cv2.imread(f, 0)
# 裁剪边缘
x, w, y, h = clip
img = img[x:w, y:h]
######### 二值处理 ##########
img_w, img_h = img.shape
# 去噪
img_med = cv2.medianBlur(img, 5)
kernel = np.zeros((7, 7), np.uint8)
thresh = cv2.morphologyEx(img_med, cv2.MORPH_OPEN, kernel)
# 不同阈值处理
# 计算均值
sums = 0
for i in range(img_w):
for j in range(img_h):
sums += thresh[i][j]
mean_value = sums // (img_w * img_h)
print()
# 计算方差的
sum_diff = 0
for i in range(img_w):
for j in range(img_h):
diff = float(
(mean_value - thresh[i][j]) * (mean_value - thresh[i][j]))
sum_diff += diff
# print(sum_diff)
variance = int((sum_diff // (img_w * img_h))**0.5)
print("mean value: ", mean_value, ", variance: ", variance)
# 计算直方图
histogram = [0 for _ in range(256)]
for i in range(img_w):
for j in range(img_h):
histogram[thresh[i][j]] += 1
# print(histogram)
for i in range(len(histogram)):
histogram[i] = str(histogram[i])
# 写入数据
# 先写入均值 直方图 方差 数据 至 data.txt
# histogram转换类型
for i in range(len(histogram)):
histogram[i] = str(histogram[i])
data = str(count) + "\t" + str(mean_value) + "\t" + str(
variance) + '\t' + "\t".join(histogram) + '\n'
data_file.write(data)
# 写入标签数据至label.txt
label = str(count) + "\t" + str(mean_value) + "\t" + str(
variance) + '\n'
label_file.write(label)
# 获取n个阈值
n = 20
gap = 2 # 两个阈值之间的间隔
thresh_value = []
small = mean_value
left = 0
while small > 0 and left < n:
thresh_value.append(small)
small -= gap
left += 1
# 大于均值的很差，所以不考虑了
large = mean_value
r = 5
right = 0
while large < 255 and right < r:
thresh_value.append(large)
large += 2
right += 1
for v in thresh_value:
ret, new_thresh = cv2.threshold(thresh, v, 255, cv2.THRESH_BINARY)
kernel = np.zeros((7, 7), np.uint8)
new_thresh = cv2.morphologyEx(new_thresh, cv2.MORPH_CLOSE, kernel)
new_thresh = cv2.medianBlur(new_thresh, 5)
# 保存
basename = os.path.basename(f)
file = os.path.splitext(basename)
file_prefix = str(count) # file[0]
file_suffix = file[-1]
if v == mean_value:
image_name = file_prefix + '_thresh_value_' + str(
v) + '_mean' + file_suffix
else:
image_name = file_prefix + '_thresh_value_' + str(
v) + file_suffix
out_file = os.path.join(out_dir, image_name)
print("saving :", out_file)
cv2.imwrite(out_file, new_thresh)
count += 1
data_file.close()
label_file.close()
record_file.close()
os.startfile(out_dir)
def process(imageDir, outputDir):
"""使用DBSACN裁剪图片"""
files = api.getFiles(imageDir)
total = len(files)
pointsDir = os.path.join(outputDir, "points")
cluserDir = os.path.join(outputDir, "cluster")
cropDir = os.path.join(outputDir, "crop")
util.mkdirs([cluserDir, cropDir, pointsDir])
marginInfo = {}
for i, f in enumerate(files):
basename = os.path.basename(f)
print("[info] crop image, process {} / {}: {}".format(
i + 1, total, basename))
img = cv2.imread(f)
cropImg = moveMargin(img, (45, -45, 45, -45)) # 删除边框特征
maxMargin = None
method = ["SIFT", "SURF", "ORB"]
mergePoints = []
for m in method:
img = copy.deepcopy(cropImg)
curpointsDir = os.path.join(pointsDir, m)
curcluserDir = os.path.join(cluserDir, m)
curcropDir = os.path.join(cropDir, m)
util.mkdirs([curcluserDir, curcropDir, curpointsDir])
points = getROIPoint(img, f=m) # 获得ROI点
# 保存ROI
pointsSavePath = os.path.join(curpointsDir, basename)
savePointsImg(pointsSavePath, points, img.shape)
try:
maxCluster = getMaxCluster(points, 40) # 获得最大聚类
except:
print("[error] skip: {}".format(basename))
continue
mergePoints.extend(maxCluster) # 融合特征点
# 保存关键点图像
clusterSavePath = os.path.join(curcluserDir, basename)
savePointsImg(clusterSavePath, maxCluster, img.shape)
margin = getMargin(maxCluster) # 获得边界点u, r, d, l
if not maxMargin:
maxMargin = margin
else:
u, r, d, l = margin
maxMargin[0] = min(maxMargin[0], u)
maxMargin[1] = max(maxMargin[1], r)
maxMargin[2] = max(maxMargin[2], d)
maxMargin[3] = min(maxMargin[3], l)
retImg = cropImage(img, margin) # 裁剪图片
# 保存裁剪结果
cropSavePath = os.path.join(curcropDir, basename)
cv2.imwrite(cropSavePath, retImg)
# 三种方法获得的最大边界
w, h = img.shape[0], img.shape[1]
if maxMargin[0] - up < 0:
maxMargin[0] = 0
else:
maxMargin[0] -= up
if maxMargin[1] + right > w:
maxMargin[1] = w - 1
else:
maxMargin[1] += right
# 图片下边缘不补偿
if maxMargin[3] - left < 0:
maxMargin[3] = 0
else:
maxMargin[3] -= left
marginInfo[basename] = maxMargin # 记录切边信息
retImg = cropImage(img, maxMargin) # 裁剪图片
# 保存不同特征点融合图像
mergeSavePath = os.path.join(pointsDir,
"mergePoints_{}".format(basename))
savePointsImg(mergeSavePath, mergePoints, img.shape)
# 保存裁剪结果
cropSavePath = os.path.join(cropDir, basename)
cv2.imwrite(cropSavePath, retImg)
# if i == 5: assert False, 'break'
writeMarginInfo(os.path.join(outputDir, "marginInfo.txt"),
marginInfo) # 保存到结果目录
os.startfile(outputDir)
def genDiffPicByThresholdValue(imgDir, outputDir): files = api.getFiles(imgDir) util.mkdirs(outputDir)
def process(imageDir, outputDir):
"""使用DBSACN裁剪图片"""
files = api.getFiles(imageDir)
total = len(files)
pointsDir = os.path.join(outputDir, "points")
cluserDir = os.path.join(outputDir, "cluster")
cropDir = os.path.join(outputDir, "crop")
util.mkdirs([cluserDir, cropDir, pointsDir])
marginInfo = {}
for i, f in enumerate(files):
basename = os.path.basename(f)
print("process {} / {}: {}".format(i+1, total, basename))
img = cv2.imread(f)
maxMargin = None
method = ["SIFT", "SURF", "ORB"]
for m in method:
curpointsDir = os.path.join(pointsDir, m)
curcluserDir = os.path.join(cluserDir, m)
curcropDir = os.path.join(cropDir, m)
util.mkdirs([curcluserDir, curcropDir, curpointsDir])
points = getROIPoint(img) # 获得ROI点
# 保存ROI
pointsSavePath = os.path.join(curpointsDir, basename)
savePointsImg(pointsSavePath, points, img.shape)
maxCluster = getMaxCluster(points, 40) # 获得最大聚类
# 保存关键点图像
clusterSavePath = os.path.join(curcluserDir, basename)
savePointsImg(clusterSavePath, maxCluster, img.shape)
margin = getMargin(maxCluster) # 获得边界点u, r, d, l
if not maxMargin:
maxMargin = margin
else:
u, r, d, l = margin
maxMargin[0] = min(maxMargin[0], u)
maxMargin[1] = max(maxMargin[1], r)
maxMargin[2] = max(maxMargin[2], d)
maxMargin[3] = min(maxMargin[3], l)
retImg = cropImage(img, margin) # 裁剪图片
# 保存裁剪结果
cropSavePath = os.path.join(curcropDir, basename)
cv2.imwrite(cropSavePath, retImg)
# 三种方法获得的最大边界
h, w = img.shape[0], img.shape[1]
print
if maxMargin[0] - up < 0:
maxMargin[0] = 0
else:
maxMargin[0] -= up
if maxMargin[1] + right > w:
maxMargin[1] = w - 1
else:
maxMargin[1] += right
if maxMargin[3] - left < 0:
maxMargin[3] = 0
else:
maxMargin[3] -= left
marginInfo[basename] = maxMargin # 记录切边信息
retImg = cropImage(img, maxMargin) # 裁剪图片
# 保存裁剪结果
cropSavePath = os.path.join(cropDir, basename)
cv2.imwrite(cropSavePath, retImg)
writeMarginInfo(os.path.join(outputDir, "marginInfo.txt"), marginInfo) # 保存到结果目录
os.startfile(outputDir)