def extract_image(self, image):
extract = ExtractImage(image)
#横向提取
left_list_draw_line = extract.horizontal_draw_line(1, 2)
#合并
left_list_split_line = extract.horizontal_merger_line(left_list_draw_line)
#纵向提取
left_list_extract_image = extract.vertical_line(1, 2, left_list_split_line)
#合并图片
list_extract_image = []
list_extract_image_path = []
for each_extract_image in left_list_extract_image:
list_extract_image.append(each_extract_image[0])
list_extract_image_path.append(each_extract_image[1])
merge_save_path = ""
if len(list_extract_image):
merge_save_path = extract.image_merge(list_extract_image)
#保存中间处理图片
deal_save_path = "static/images/dealimg/" + time.ctime() + ".jpg"
extract.image.save("../%s" % deal_save_path)
list_save_path = []
list_save_path.append([deal_save_path, merge_save_path, list_extract_image_path])
return list_save_path
def extract_image(self):
#读取数据集urlno，获取专利号
results = self.db.urlno.find({"kind": "patent", "downflag": 1, "extractflag": 0}, {"_id": 0, "indexflag": 1, "zippath": 1})
if results:
patent_handler = PatentClass()
list_patentno = []
list_zip_path = []
for result in results:
list_patentno.append(result["indexflag"])
list_zip_path.append(result["zippath"])
#解压文件
list_unzip_save_path = []
for patentno, zip_path in zip(list_patentno, list_zip_path):
list_unzip_save_path = patent_handler.unzip_patent(zip_path)
print "解压完成"
#判断是否解压成功
if len(list_unzip_save_path):
#反向提取图片
list_len = len(list_unzip_save_path)
count_none = 0
list_deal_save_path = []
list_merge_save_path = []
list_extract_save_path = []
for i in xrange(list_len):
image_path = list_unzip_save_path[list_len-i-1]
#将tif暂时转成jpg
PythonMagick.Image("../%s" % image_path).write("temptiftojpg.jpg")
image = Image.open("temptiftojpg.jpg")
image = image.convert("L")
#提取图片
list_save_path = patent_handler.extract_image(image)
if len(list_save_path[0][1]):
list_deal_save_path.append(list_save_path[0][0])
list_merge_save_path.append(list_save_path[0][1])
list_extract_save_path.append(list_save_path[0][2])
else:
count_none += 1
if count_none == 2:
break
#将多张图片合成一张大图
final_merge_save_path = ""
if len(list_merge_save_path):
list_merge_image = []
for merge_image_path in list_merge_save_path:
image = Image.open("../%s" % merge_image_path)
list_merge_image.append(image)
patent_num = patentno
extract = ExtractImage()
final_merge_save_path = extract.image_merge(list_merge_image, patentno)
#存入数据集，更新数据集urlno
try:
self.db.urlno.update({"indexflag": patentno}, {"$set": {"extractflag": 1, "dealpath": list_deal_save_path,
"extractpath": list_extract_save_path, "mergepath": final_merge_save_path}})
print "提取成功"
except: pass
return
def extract_image(self):
#读取数据集pdfurl中图片提取标志为0的url，并返回url
list_url = self.db.urlno.find({"convertflag": 1, "extractflag": 0}, {"indexflag": 1})
i = 0
for urls in list_url:
url = urls["indexflag"]
#读取数据集convertimg中url想对应的转换图片
list_convert_img = self.db.convertimg.find({"indexflag": url}, {"convertpath": 1})
for item in list_convert_img:
list_convert_imgs = item["convertpath"]
#提取图片
list_save_path = []
for convert_img in list_convert_imgs:
image = Image.open(convert_img)
image = image.convert("L")
extract_img = ExtractImage(image)
list_each_save_path = extract_img.main()
list_save_path.append(list_each_save_path)
#分别提取不同路径
list_deal_image_path = []
list_extract_image_path = []
list_merge_image_path = []
for each_save_path in list_save_path:
list_deal_image_path.append(each_save_path[0][0])
if len(each_save_path[0][1]):
list_merge_image_path.append(each_save_path[0][1])
if len(each_save_path[0][2]):
for each_extract_image_path in each_save_path[0][2]:
list_extract_image_path.append(each_extract_image_path)
#合并单页图片
final_merge_image_path = ""
if len(list_merge_image_path):
list_merge_image = []
for each_merge_image_path in list_merge_image_path:
image = Image.open(each_merge_image_path)
list_merge_image.append(image)
extract_img = ExtractImage()
final_merge_image_path = extract_img.image_merge(list_merge_image)
#写入数据集dealimg,extractimg,mergeimg
exist_result = self.db.dealimg.find_one({"indexflag": url})
if not exist_result:
self.db.dealimg.insert({"indexflag": url, "dealpath": list_deal_image_path, "time": time.ctime()})
exist_result = self.db.extractimg.find_one({"indexflag": url})
if not exist_result:
self.db.extractimg.insert({"indexflag": url, "extractpath": list_extract_image_path, "time": time.ctime()})
exist_result = self.db.mergeimg.find_one({"indexflag": url})
if not exist_result:
self.db.mergeimg.insert({"indexflag": url, "mergepath": final_merge_image_path, "time": time.ctime()})
self.db.urlno.update({"indexflag": url}, {"$set": {"extractflag": 1}})
i = i + 1
print i
if i > 9:
break
return