PDF&图片中文字采集处理踩坑记录

1.最好用的是e-iceblue公司的 spire.pdf For java 免费版就够用了
可以自动PDF中的文字和图片

<dependency>
	<groupId>e-iceblue</groupId>
	<artifactId>spire.pdf.free</artifactId>
	<version>2.2.2</version>
</dependency>

2.org.apache.pdfbox
将PDF的每一页转成一张图片

<dependency>
	<groupId>org.apache.pdfbox</groupId>
	<artifactId>pdfbox</artifactId>
	<version>2.0.6</version>
</dependency>

3.net.sourceforge.tess4j
最基本的OCR识别工具，可以体验驯宠的乐趣

<dependency>
	<groupId>net.sourceforge.tess4j</groupId>
	<artifactId>tess4j</artifactId>
	<version>3.4.8</version>
</dependency>

@ResponseBody
	@RequestMapping(value = "imageUpload", method = RequestMethod.POST)
    public Map<String, Object> springUpload(@RequestParam(value="file", required=false)List<MultipartFile> files,
    		HttpServletRequest request) throws IllegalStateException, IOException//, TesseractException
    {
		Map<String, Object> msgmap = new HashMap<String, Object>();
		for(MultipartFile file:files){
			if(file != null){
				
				//设置文件的保存路径
				CommonUtils.delFile(AddressUtil.getUpload_address()+ "\\"+"图片识别"+"\\",file.getOriginalFilename());
				String filePath = AddressUtil.getUpload_address() + "\\"+"图片识别"+"\\"+ file.getOriginalFilename();
	            CommonUtils.fileMake(filePath);  
	            //转存文件
	            try {
	            	file.transferTo(new File(filePath));
	            	msgmap.put("message", "success");
	            } catch (IOException e) {
	            	msgmap.put("message", "error");
	                e.printStackTrace();
	            }
	            
	            //获取上传文件的扩展名
				String suffix = file.getOriginalFilename().substring(file.getOriginalFilename().lastIndexOf(".") + 1);
				//解析PDF
				if("PDF".equals(suffix.toUpperCase())){
					
					/** org.apache.pdfbox 将PDF逐页转成图片处理 一页内容放到一张图片里  
					// 转换前文件地址
					File scrFile = new File(AddressUtil.getUpload_address()+ "\\"+"图片识别"+"\\",file.getOriginalFilename());
					try {
						//转换前PDF本尊
						PDDocument doc = PDDocument.load(scrFile);
						//调用转换器
						PDFRenderer render = new PDFRenderer(doc);
						//PDF页数取得
						int pageCount = doc.getNumberOfPages();
						for (int i = 0; i < pageCount; i++) {
							//写入图片buffer
							BufferedImage image = render.renderImageWithDPI(i, 144); // Windows native DPI
							//转换后图片 Created
							ImageIO.write(image, "PNG", new File(AddressUtil.getUpload_address()+ "\\"+"图片识别"+"\\",file.getOriginalFilename()+"_"+(i+1)+".png"));
							//调用图片识别接口>...
							msgmap.put("foundText", postHttp(AddressUtil.getUpload_address() + "\\"+"图片识别"+"\\" + file.getOriginalFilename()+"_"+(i+1)+".png"));
						}
					} catch (IOException e) {
						e.printStackTrace();
					}
					*/
					try {
						//转换前PDF本尊
						PdfDocument doc = new PdfDocument();
						doc.loadFromFile(
								AddressUtil.getUpload_address() + "\\" + "图片识别" + "\\" + file.getOriginalFilename());
						//PDF中的文字 Container
						StringBuilder sb = new StringBuilder();
						//PDF文件的页对象
						PdfPageBase page;
						for (int i = 0; i < doc.getPages().getCount(); i++) {
							//页对象取得
							page = doc.getPages().get(i);
							//提取文字部分
							sb.append(page.extractText(true));
							//如果有图片则提取
							if (page.extractImages() != null) {
								int count = 1;
								for (BufferedImage image : page.extractImages()) {
									if (image != null) {
										//指定输出图片名
										File output = new File(AddressUtil.getUpload_address() + "\\" + "图片识别" + "\\",
												file.getOriginalFilename() + "_" + count + ".png");
										//图片转存
										ImageIO.write(image, "PNG", output);
										
										//调用识别图片接口
										msgmap.put("foundInfo", postHttp(AddressUtil.getUpload_address() + "\\" + "图片识别"
												+ "\\" + file.getOriginalFilename() + "_" + (i + 1) + ".png"));
									}
									count = count + 1;
								}
							}
						}
						System.out.println(sb);
						doc.close();
					} catch (Exception e) {
						e.printStackTrace();
					}
				//解析图片	
				}else {
		            /** net.sourceforge.tess4j 图片识别度低 需驯化生成需要的字库
					ITesseract instance = new Tesseract();
		            
		            //如果需要识别英文之外的语种，需要指定识别语种，并且需要将对应的语言包放进项目中
		            instance.setLanguage("chi_sim");
		            
		            // 指定识别图片
		            File imgDir = new File(AddressUtil.getUpload_address()+ "\\"+"图片识别"+"\\",file.getOriginalFilename());
		            long startTime = System.currentTimeMillis();
		            String ocrResult = instance.doOCR(imgDir);
		            System.out.println("OCR Result: \n" + ocrResult + "\n 耗时：" + (System.currentTimeMillis() - startTime) + "ms");
		            */
					//调用识别图片接口
					msgmap.put("foundInfo", postHttp(AddressUtil.getUpload_address() + "\\"+"图片识别"+"\\" + file.getOriginalFilename()));
				}
	        }
		}
		return msgmap;
    }