目录
首先安装Tesseract-OCR
python实现获取
JAVA实现获取
首先安装Tesseract-OCR
可自行搜索Tesseract-OCR安装方法
python实现获取
提高识别成功率：
# 转换为灰度图像 image = image.convert('L') image.save('L.png')# 二值化处理 threshold = 130 table = [] for i in range(256):if i < threshold:table.append(0)else:table.append(1) image = image.point(table, '1')需要优化：
请求失败，可以根据返回的结果。判断重新获取验证码，还是返回响应失败的原因。
# Tesseract-OCR 读取干扰验证码
import time
from PIL import Image
import pytesseract
import requestsdef searchId(waybillNo):# 获取会话中的JSESSIONID------response = requests.get('https://www.xxx.com/api/verifyCode?' + str(int(time.time() * 1000)))# 获取响应中的 cookiejession_id = response.cookies.get('JSESSIONID') # 打印 cookieprint(jession_id)try:# 获取验证码CodeUrl = 'https://www.xxx.com/api/verifyCode?' + str(int(time.time() * 1000))print(CodeUrl)headers = {"Cookie": "_bl_uid=7Omv727F2vUlFwjLtaXkav10ggsa; JSESSIONID="+jession_id}# 携带Cookie发送请求获取验证码图片response = requests.get(CodeUrl,headers=headers)# 保存图片if response.status_code == 200:with open("verify_code.png", "wb") as f:f.write(response.content)print("验证码图片已保存为 verify_code.png")else:print("请求失败，状态码:", response.status_code)# 打开图片并进行处理image = Image.open('verify_code.png')# 转换为灰度图像image = image.convert('L')image.save('L.png')# 二值化处理threshold = 130table = []for i in range(256):if i < threshold:table.append(0)else:table.append(1)image = image.point(table, '1')# 查看处理后的图片# image.save('2.png')# 获取图片数字codeText = pytesseract.image_to_string(image, config='--psm 10 --oem 3 -c tessedit_char_whitelist=0123456789')# 携带运单号+验证码+cookie获取信息BASE_URL = "https://www.xxx.com" # 请替换为实际的BASE_URL#waybillNo = "93387172455" # 请替换为实际的运单号# 拼接请求获取数据dhurl = f"{BASE_URL}/eos/awb/{waybillNo}/{codeText}"print(dhurl.replace('\n',''))response = requests.get(dhurl.replace('\n',''), headers=headers)data = response.json()print(data)except Exception as e:print("获取失败:" + waybillNo)if __name__ == '__main__':searchId('93387172455')JAVA实现获取
@PostMapping(value = "/getPdInfo", produces = "text/html;charset=UTF-8")@ResponseBodypublic String getPdInfo(@RequestBody List<String> items) {StringBuilder sb = new StringBuilder();for (String item : items) {if (!item.equals("")) {sb.append(getPdDhHtml(item)+"<hr>");}}return sb.toString();}private static JsonNode getDhJson(String dh) {String jsessionid="";try {// 获取会话 中的JSESSIONID// 创建 URL 对象URL url1 = new URL("https://www.xxx.com/api/verifyCode?" + System.currentTimeMillis());// 打开连接HttpURLConnection connection1 = (HttpURLConnection) url1.openConnection();// 设置请求方法connection1.setRequestMethod("GET");// 获取响应码int responseCode1 = connection1.getResponseCode();//System.out.println("Response Code: " + responseCode1);// 获取 Set-CookieString cookies = connection1.getHeaderField("Set-Cookie");// 使用正则表达式匹配 JSESSIONID 的值Pattern pattern = Pattern.compile("JSESSIONID=([^;]+);");Matcher matcher = pattern.matcher(cookies);// String jsessionid="";if (matcher.find()) {jsessionid = matcher.group(1);} else {System.out.println("未找到 JSESSIONID");}// 关闭连接connection1.disconnect();// 获取验证码String codeUrl = "https://www.xxx.com/api/verifyCode?" + System.currentTimeMillis();Map<String, String> headers = new HashMap<>();headers.put("Cookie", "_bl_uid=7Omv727F2vUlFwjLtaXkav10ggsa; JSESSIONID="+jsessionid);URL url = new URL(codeUrl);HttpURLConnection connection = (HttpURLConnection) url.openConnection();connection.setRequestMethod("GET");for (Map.Entry<String, String> entry : headers.entrySet()) {connection.setRequestProperty(entry.getKey(), entry.getValue());}int responseCode = connection.getResponseCode();if (responseCode == 200) {InputStream inputStream = connection.getInputStream();FileOutputStream fileOutputStream = new FileOutputStream(jsessionid+"verify_code.png");byte[] buffer = new byte[1024];int bytesRead;while ((bytesRead = inputStream.read(buffer)) != -1) {fileOutputStream.write(buffer, 0, bytesRead);}fileOutputStream.close();inputStream.close();} else {System.out.println("请求失败，状态码: " + responseCode);}BufferedImage image = ImageIO.read(new File(jsessionid+"verify_code.png"));// 转换为灰度图像BufferedImage grayImage = new BufferedImage(image.getWidth(), image.getHeight(), BufferedImage.TYPE_BYTE_GRAY);grayImage.getGraphics().drawImage(image, 0, 0, null);// 保存灰度图像文件
// try {
// File output = new File("path_to_save_gray_image.png");
// ImageIO.write(grayImage, "png", output);
// System.out.println("灰度图像已保存到: " + output.getAbsolutePath());
// } catch (IOException e) {
// e.printStackTrace();
// }// 二值化处理int threshold = 180;for (int y = 0; y < grayImage.getHeight(); y++) {for (int x = 0; x < grayImage.getWidth(); x++) {int rgb = grayImage.getRGB(x, y);int gray = (rgb >> 16) & 0xFF;if (gray < threshold) {grayImage.setRGB(x, y, 0xFF000000); // 黑色} else {grayImage.setRGB(x, y, 0xFFFFFFFF); // 白色}}}// 保存二值化图像到文件
// try {
// File output = new File("path_to_save_binarized_image.png");
// ImageIO.write(grayImage, "png", output);
// System.out.println("二值化图像已保存到: " + output.getAbsolutePath());
// } catch (IOException e) {
// e.printStackTrace();
// }// 获取图片数字Tesseract tesseract = new Tesseract();tesseract.setDatapath("D:\\work\\Tesseract-OCR\\tessdata"); // 设置tessdata路径tesseract.setLanguage("eng"); // 设置语言为英语tesseract.setTessVariable("tessedit_char_whitelist", "0123456789"); // 设置字符白名单为数字String codeText = tesseract.doOCR(grayImage);String baseUrl = "https://www.xxx.com";// String waybillNo = "93387172455";String dhurl = baseUrl + "/eos/awb/" + dh + "/" + codeText.replaceAll("\\D","");URL dhUrl = new URL(dhurl);HttpURLConnection dhConnection = (HttpURLConnection) dhUrl.openConnection();dhConnection.setRequestMethod("GET");for (Map.Entry<String, String> entry : headers.entrySet()) {dhConnection.setRequestProperty(entry.getKey(), entry.getValue());}int dhResponseCode = dhConnection.getResponseCode();if (dhResponseCode == 200) {BufferedReader in = new BufferedReader(new InputStreamReader(dhConnection.getInputStream()));String inputLine;StringBuffer response = new StringBuffer();while ((inputLine = in.readLine()) != null) {response.append(inputLine);}in.close();
// System.out.println(response.toString());ObjectMapper objectMapper = new ObjectMapper();JsonNode rootNode = objectMapper.readTree(response.toString());String msg = rootNode.get("msg").asText();if ("null".equals(msg)) {//等于null说明成功return rootNode.get("data");}else if ("运单号不合法！！".equals(msg)){return rootNode.get("msg");}else if ("系统异常".equals(msg)){return rootNode.get("msg");}} else {System.out.println("请求失败，状态码: " + dhResponseCode);}} catch (Exception e) {System.out.println("异常"+e);}finally {//删除验证码图片File imageFile = new File(jsessionid+"verify_code.png");if (imageFile.exists()) {imageFile.delete();
// boolean isDeleted = imageFile.delete();
// if (isDeleted) {
// System.out.println("图片已成功删除");
// } else {
// System.out.println("图片删除失败");
// }}}// 失败则再次发起return getDhJson(dh);}