init envs

README.md

@@ -0,0 +1,47 @@
+## 本项目基于[yolo3](https://github.com/pjreddie/darknet.git) 与[crnn](https://github.com/meijieru/crnn.pytorch.git)  实现中文自然场景文字检测及识别
+
+## 环境部署
+python=3.6 pytorch=0.2.0
+``` Bash
+git clone https://github.com/pjreddie/darknet.git
+cd darknet
+make 
+cd ..
+git clone https://github.com/chineseocr/chineseocr.git
+cd chineseocr
+sh setup.sh(cpu sh setpu-cpu.sh)
+```
+
+## 下载模型文件   
+
+将近期更新
+
+``` Bash
+mv models.zip  chineseocr
+unzip models.zip 
+```
+## web服务启动
+
+``` Bash
+cd chineseocr## 进入chineseocr目录
+ipython app.py 8080 ##8080端口号，可以设置任意端口
+```
+
+## 识别结果展示
+<div>
+<img width="300" height="300" src="https://github.com/chineseocr/chinsesocr/blob/master/test/img1.png"/>
+<img width="300" height="300" src="https://github.com/chineseocr/chinsesocr/blob/master/test/4.png"/>
+</div>
+
+## 访问服务
+http://127.0.0.1:8080/ocr
+
+<img width="300" height="300" src="https://github.com/chineseocr/chinsesocr/blob/master/test/demo.png"/>
+
+
+## 参考
+1. yolo3 https://github.com/pjreddie/darknet.git   
+2. crnn  https://github.com/meijieru/crnn.pytorch.git              
+3. ctpn  https://github.com/eragonruan/text-detection-ctpn    
+4. CTPN  https://github.com/tianzhi0549/CTPN     
+

app.py

@@ -0,0 +1,74 @@
+# -*- coding: utf-8 -*-
+"""
+@author: lywen
+"""
+import os
+from PIL import Image
+import json
+import time
+import uuid
+import numpy as np
+import sys
+import base64
+
+import web  
+
+web.config.debug  = True    
+from apphelper.image import convert_image,read_url_img,string_to_array,array_to_string,base64_to_array
+import model
+render = web.template.render('templates', base='base')
+
+
+class OCR:
+    """通用OCR识别"""
+
+    def GET(self):  
+        post = {}
+        post['postName'] = u'ocr'##请求地址
+        post['height'] = 1000
+        post['H'] = 1000
+        post['width'] = 600
+        post['W'] = 600
+        post['uuid'] = uuid.uuid1().__str__()
+        return render.ocr(post)
+
+    def POST(self): 
+        data = web.data()
+        data = json.loads(data)
+        imgString = data['imgString'].encode().split(b';base64,')[-1]
+        imgString = base64.b64decode(imgString)
+        jobid = uuid.uuid1().__str__()
+        path = '/tmp/{}.jpg'.format(jobid)
+        with open(path,'wb') as f:
+            f.write(imgString)
+        img = Image.open(path).convert("RGB")
+        W,H = img.size
+        _,result,angle= model.model(img,detectAngle=True,config=dict(MAX_HORIZONTAL_GAP=200,           
+                MIN_V_OVERLAPS=0.6,
+                MIN_SIZE_SIM=0.6,
+                TEXT_PROPOSALS_MIN_SCORE=0.2,
+                TEXT_PROPOSALS_NMS_THRESH=0.3,
+                TEXT_LINE_NMS_THRESH = 0.99,
+                MIN_RATIO=1.0,
+                LINE_MIN_SCORE=0.2,
+                TEXT_PROPOSALS_WIDTH=5,
+                MIN_NUM_PROPOSALS=0,
+                ),
+                leftAdjust=True,rightAdjust=True,alph=0.1)
+
+        res = map(lambda x:{'w':x['w'],'h':x['h'],'cx':x['cx'],'cy':x['cy'],'degree':x['degree'],'text':x['text']}, result) 
+        res = list(res)
+
+        os.remove(path)
+        return json.dumps(res,ensure_ascii=False)
+
+
+
+
+urls = (u'/ocr',u'OCR',
+       )
+
+if __name__ == "__main__":  
+
+      app = web.application(urls, globals())   
+      app.run()

apphelper/__init__.py

@@ -0,0 +1,41 @@
+# -*- coding: utf-8 -*-
+"""
+@author: lywen
+todo:通用函数文件
+"""
+import datetime as dt
+
+def get_date():
+    """
+    获取当前时间
+    """
+    try:
+        now = dt.datetime.now()
+        nowString = now.strftime('%Y%m%d')
+    except:
+        nowString = '000000'
+    return nowString
+
+def strdate_to_date(string,format='%Y-%m-%d %H:%M:%S'):
+    try:
+        return dt.datetime.strptime(string,format)
+    except:
+        return dt.datetime.now()
+
+def diff_time(beginDate,endDate,format='%Y-%m-%d %H:%M:%S'):
+    str1Date = strdate_to_date(beginDate,format)
+    str2Date = strdate_to_date(endDate,format)
+    times    = str2Date - str1Date
+    return   times.total_seconds()
+
+def get_now():
+    """
+    获取当前时间
+    """
+    try:
+        now = dt.datetime.now()
+        nowString = now.strftime('%Y-%m-%d %H:%M:%S')
+    except:
+        nowString = '00-00-00 00:00:00'
+    return nowString
+

apphelper/image.py

@@ -0,0 +1,104 @@
+# -*- coding: utf-8 -*-
+"""
+##图像相关函数
+@author: lywen
+"""
+import base64
+#import urllib2
+import requests
+import numpy as np
+import cv2
+from PIL import Image
+import sys
+import six
+import traceback
+import uuid
+
+
+
+from PIL import Image   
+def read_img(path):
+    im = Image.open(path).convert('RGB')
+    img = np.array(im)
+    return img 
+
+def convert_image(path=None,string=None):  
+    # Picture ==> base64 encode 
+    if path is not None:
+        with open(path, 'rb') as f:  
+            base64_data = base64.b64encode(f.read())
+        return base64_data
+    if string is not None:
+        base64_data = base64.b64encode(string)
+        return base64_data
+
+def string_to_array(string):
+    if check_image_is_valid(string):
+                buf = six.BytesIO()
+                buf.write(string)
+                buf.seek(0)
+                img = np.array(Image.open(buf))
+                return img
+    else:
+            return None
+
+def check_image_is_valid(imageBin):
+    """
+    检查图片是否有效
+    """
+    if imageBin is None:
+        return False
+    imageBuf = np.fromstring(imageBin, dtype=np.uint8)
+
+    img = cv2.imdecode(imageBuf, cv2.IMREAD_GRAYSCALE)
+    if img is None:
+        return False
+
+    imgH, imgW = img.shape[0], img.shape[1]
+    if imgH * imgW == 0:
+        return False
+    return True
+
+def base64_to_array(string):
+    #try:
+            base64_data = base64.b64decode(string)
+            buf = six.BytesIO()
+            buf.write(string)
+            buf.seek(0)
+            img = np.array(Image.open(buf))
+            return img
+
+    #except:
+     #   return None
+
+
+
+def read_url_img(url):
+    """
+    爬取网页图片
+    """
+    try:
+        req = requests.get(url,timeout=5)##访问时间超过5s，则超时
+        if req.status_code==200:
+            imgString = req.content
+        #imgString = req.read()
+        if check_image_is_valid(imgString):
+                buf = six.BytesIO()
+                buf.write(imgString)
+                buf.seek(0)
+                img = Image.open(buf).convert('RGB')
+                return img
+        else:
+            return None
+    except:
+        traceback.print_exc()
+        return None
+
+
+def array_to_string(array):
+    image = Image.fromarray(array)
+    output = six.BytesIO()
+    image.save(output,format='png')
+    contents = output.getvalue()
+    output.close()
+    return contents

config.py

@@ -0,0 +1,7 @@
+import os
+darknetRoot = os.path.join("..","darknet")## yolo 安装目录
+pwd = os.getcwd()
+yoloCfg = os.path.join(pwd,"models","text.cfg")
+yoloWeights = os.path.join(pwd,"models","text.weights")
+yoloData = os.path.join(pwd,"models","text.data")
+ocrModel = os.path.join(pwd,"models","ocr.pth")

crnn/crnn.py

@@ -0,0 +1,69 @@
+#coding:utf-8
+import sys
+sys.path.insert(1, "./crnn")
+import torch
+import torch.utils.data
+from torch.autograd import Variable 
+from crnn import util
+from crnn import dataset
+from crnn.models import crnn as crnn
+from crnn import keys
+#from conf import crnnModelPath
+#from conf import GPU
+GPU=False
+from collections import OrderedDict
+from config import ocrModel
+def crnnSource():
+    alphabet = keys.alphabet
+    converter = util.strLabelConverter(alphabet)
+    if torch.cuda.is_available() and GPU:
+       model = crnn.CRNN(32, 1, len(alphabet)+1, 256, 1).cuda()
+    else:
+        model = crnn.CRNN(32, 1, len(alphabet)+1, 256, 1).cpu()
+
+    state_dict = torch.load(ocrModel)
+    new_state_dict = OrderedDict()
+    for k, v in state_dict.items():
+        name = k.replace('module.','') # remove `module.`
+        new_state_dict[name] = v
+    # load params
+
+    model.load_state_dict(new_state_dict)
+    model.eval()
+
+    return model,converter
+
+##加载模型
+model,converter = crnnSource()
+
+def crnnOcr(image):
+       """
+       crnn模型，ocr识别
+       @@model,
+       @@converter,
+       @@im
+       @@text_recs:text box
+
+       """
+       scale = image.size[1]*1.0 / 32
+       w = image.size[0] / scale
+       w = int(w)
+       #print "im size:{},{}".format(image.size,w)
+       transformer = dataset.resizeNormalize((w, 32))
+       if torch.cuda.is_available() and GPU:
+           image = transformer(image).cuda()
+       else:
+           image = transformer(image).cpu()
+
+       image = image.view(1, *image.size())
+       image = Variable(image)
+       model.eval()
+       preds = model(image)
+       _, preds = preds.max(2)
+       preds = preds.transpose(1, 0).contiguous().view(-1)
+       preds_size = Variable(torch.IntTensor([preds.size(0)]))
+       sim_pred = converter.decode(preds.data, preds_size.data, raw=False)
+
+       return sim_pred
+
+