Esempio n. 1
0
def main(fn):
    # 读取并预处理验证码
    img = cv2.imread(fn, cv2.IMREAD_GRAYSCALE)
    text = pretreatment.get_text(img)
    imgs = pretreatment.get_imgs(img)
    text = text / 255.0
    h, w = text.shape
    text.shape = (1, h, w, 1)

    # 识别文字
    model = models.load_model('model.h5')
    label = model.predict(text)
    label = label.argmax()
    print(label)

    # 加载图片分类器
    data = np.load('images.npz')
    images, labels = data['images'], data['labels']
    labels = labels.argmax(axis=1)
    for pos, img in enumerate(imgs):
        try:
            img.dtype = np.uint64
            img = img[0]
            idx = list(images).index(img)
            label = labels[idx]
            print(pos // 4, pos % 4, label)
        except:
            print('unknown')
Esempio n. 2
0
def get_text(img, offset=0):
    text = pretreatment.get_text(img, offset)
    text = cv2.cvtColor(text, cv2.COLOR_BGR2GRAY)
    text = text / 255.0
    h, w = text.shape
    text.shape = (1, h, w, 1)
    return text
Esempio n. 3
0
def main(fn):
    # 读取并预处理验证码
    img = cv2.imread(fn)
    text = pretreatment.get_text(img)
    text = cv2.cvtColor(text, cv2.COLOR_BGR2GRAY)
    imgs = np.array(list(pretreatment._get_imgs(img)))
    imgs = imgs / 255.0
    text = text / 255.0
    h, w = text.shape
    text.shape = (1, h, w, 1)
    _, h, w, _ = imgs.shape
    imgs.shape = (-1, h, w, 3)

    # 识别文字
    model = models.load_model('model.h5')
    label = model.predict(text)
    label = label.argmax()
    print(label)

    # 加载图片分类器
    model = models.load_model('12306.image.model.h5')
    labels = model.predict(imgs)
    labels = labels.argmax(axis=1)
    for pos, label in enumerate(labels):
        print(pos // 4, pos % 4, label)
Esempio n. 4
0
 def get_text(self, img):
     img_array = np.array(img)
     text = pretreatment.get_text(img_array)
     text = cv2.cvtColor(text, cv2.COLOR_BGR2GRAY)
     text = text / 255.0
     h, w = text.shape
     text.shape = (1, h, w, 1)
     return text
Esempio n. 5
0
def get_text(img, offset=0):
    """
    获取图片文本信息
    :param img: 验证码图片句柄,cv2.imread返回值
    :param offset:
    :return: 图片处理后的文本信息
    """
    text = pretreatment.get_text(img, offset)
    text = cv2.cvtColor(text, cv2.COLOR_BGR2GRAY)
    text = text / 255.0
    h, w = text.shape
    text.shape = (1, h, w, 1)
    return text
Esempio n. 6
0
    def verify(self, i):
        verify_titles = [
            '打字机', '调色板', '跑步机', '毛线', '老虎', '安全帽', '沙包', '盘子', '本子', '药片',
            '双面胶', '龙舟', '红酒', '拖把', '卷尺', '海苔', '红豆', '黑板', '热水袋', '烛台', '钟表',
            '路灯', '沙拉', '海报', '公交卡', '樱桃', '创可贴', '牌坊', '苍蝇拍', '高压锅', '电线',
            '网球拍', '海鸥', '风铃', '订书机', '冰箱', '话梅', '排风机', '锅铲', '绿豆', '航母',
            '电子秤', '红枣', '金字塔', '鞭炮', '菠萝', '开瓶器', '电饭煲', '仪表盘', '棉棒', '篮球',
            '狮子', '蚂蚁', '蜡烛', '茶盅', '印章', '茶几', '啤酒', '档案袋', '挂钟', '刺绣', '铃铛',
            '护腕', '手掌印', '锦旗', '文具盒', '辣椒酱', '耳塞', '中国结', '蜥蜴', '剪纸', '漏斗',
            '锣', '蒸笼', '珊瑚', '雨靴', '薯条', '蜜蜂', '日历', '口哨'
        ]
        img = Image.open("../pic/verify{}.png".format(i))
        text = self.get_text(img)
        imgs = np.array(list(pretreatment._get_imgs(img)))
        imgs = preprocess_input(imgs)
        text_list = []
        self.LoadTextModel()
        label = self.textModel.predict(text)
        label = label.argmax()
        text = verify_titles[label]
        text_list.append(text)
        print("题目是{}".format(text))

        # 获取下一个词
        # 根据第一个词的长度来定位第二个词的位置
        if len(text) == 1:
            offset = 27
        elif len(text) == 2:
            offset = 47
        else:
            offset = 60
        text = pretreatment.get_text(img, offset=offset)
        if text.mean() < 0.95:
            label = self.textModel.predict(text)
            label = label.argmax()
            text = verify_titles[label]
            text_list.append(text)
        print("题目是{}".format(text_list))

        # 加载图片分类器
        self.LoadImgModel()
        labels = self.imgModel.predict(imgs)
        labels = labels.argmax(axis=1)
        results = []
        for pos, label in enumerate(labels):
            l = verify_titles[label]
            print(pos + 1, l)
            if l in text_list:
                results.append(str(pos + 1))
        print(results)
        return results
Esempio n. 7
0
import os

import cv2
import numpy as np

import pretreatment


path = 'imgs'
texts = []
images = []
i = 0
for fn in os.listdir(path):
    i += 1
    fn = os.path.join(path, fn)
    im = cv2.imread(fn)
    text = pretreatment.get_text(im)
    texts.append(text)
    images.append(list(pretreatment._get_imgs(im)))
    if i >= 2000:
        break
np.savez_compressed('dataset.npz', texts=texts, images=images)
Esempio n. 8
0
# coding: utf-8
import sys
 
import cv2
import numpy as np
from keras import models
 
import pretreatment
from mlearn_for_image import preprocess_input
 
 
def get_text(img, offset=0):
    text = pretreatment.get_text(img, offset)
    text = cv2.cvtColor(text, cv2.COLOR_BGR2GRAY)
    text = text / 255.0
    h, w = text.shape
    text.shape = (1, h, w, 1)
    return text
 
 
def main(fn):
    # 读取并预处理验证码
    img = cv2.imread(fn)
    text = get_text(img)
    imgs = np.array(list(pretreatment._get_imgs(img)))
    imgs = preprocess_input(imgs)
 
    # 识别文字
    model = models.load_model('model.v2.0.h5')
    label = model.predict(text)
    label = label.argmax()