# feat:使用RapidOCR4j进行OCR识别;

- 解决使用JNI导致堆外内存占用过大的问题

pom.xml

@@ -69,19 +69,19 @@
         </dependency>
         <!--  rapidocr这个必须要引入     -->
         <!--  可前往maven中央仓库https://central.sonatype.com/artifact/io.github.mymonstercat/rapidocr/versions，查看版本      -->
-        <dependency>
-            <groupId>space.anyi</groupId>
-            <artifactId>rapidocr</artifactId>
-            <version>0.0.7</version>
-        </dependency>
+        <!--<dependency>-->
+        <!--    <groupId>space.anyi</groupId>-->
+        <!--    <artifactId>rapidocr</artifactId>-->
+        <!--    <version>0.0.7</version>-->
+        <!--</dependency>-->
 
         <!--  一般只需要引入一个，CPU端建议使用onnx，移动端建议使用ncnn     -->
         <!--  可前往maven中央仓库https://central.sonatype.com/artifact/io.github.mymonstercat/rapidocr-onnx-platform/versions，查看版本      -->
-        <dependency>
-            <groupId>space.anyi</groupId>
-            <artifactId>rapidocr-onnx-platform</artifactId>
-            <version>0.0.7</version>
-        </dependency>
+        <!--<dependency>-->
+        <!--    <groupId>space.anyi</groupId>-->
+        <!--    <artifactId>rapidocr-onnx-platform</artifactId>-->
+        <!--    <version>0.0.7</version>-->
+        <!--</dependency>-->
 
         <!--<dependency>-->
         <!--    <groupId>io.github.mymonstercat</groupId>-->
@@ -94,6 +94,13 @@
             <version>5.12.2</version>
             <scope>test</scope>
         </dependency>
+
+        <dependency>
+            <groupId>io.github.hzkitty</groupId>
+            <artifactId>rapidocr4j</artifactId>
+            <version>1.0.3</version>
+        </dependency>
+
     </dependencies>
 
     <properties>

src/main/java/space/anyi/amountVerification/App.java

@@ -4,8 +4,7 @@ import com.alibaba.excel.EasyExcel;
 import com.alibaba.excel.ExcelWriter;
 import com.alibaba.excel.read.listener.PageReadListener;
 import com.alibaba.excel.write.metadata.WriteSheet;
-import io.github.mymonstercat.Model;
-import io.github.mymonstercat.ocr.InferenceEngine;
+import io.github.hzkitty.RapidOCR;
 import javafx.application.Application;
 import javafx.application.Platform;
 import javafx.geometry.Insets;
@@ -28,6 +27,7 @@ import space.anyi.amountVerification.model.InputModel;
 import space.anyi.amountVerification.model.OutputErrorModel;
 import space.anyi.amountVerification.model.OutputModel;
 
+import javax.imageio.ImageIO;
 import java.awt.image.BufferedImage;
 import java.io.File;
 import java.io.IOException;
@@ -98,15 +98,13 @@ public class App extends Application {
 
     /**
      * OCR使用的对象
-     * @see InferenceEngine
+     * @see RapidOCR
      */
-    private final InferenceEngine engineOCR;
+    private final RapidOCR rapidOCR;
 
     public App() {
-        //配置临时目录,用于存放OCR模型
-        System.setProperty("java.io.tmpdir", System.getenv("user.dir") + "\\temp");
         //初始化OCR使用的对象
-        this.engineOCR = InferenceEngine.getInstance(Model.ONNX_PPOCR_V4);
+        this.rapidOCR = new RapidOCR();
         //初始化匹配金额的正则表达式对象
         String regex = "理赔金额合计[^\\d]{1,2}[￥¥#Y半]?(?<money>\\d+\\.\\d{2})";
         this.pattern = Pattern.compile(regex);
@@ -464,11 +462,10 @@ public class App extends Application {
         BufferedImage bufferedImage = null;
         try {
             //使用RapidOCR进行识别
-            text = engineOCR.runOcr(workOrderFile.getAbsolutePath()).getStrRes().trim();
             //开始识别整张图片中的文字
-            //bufferedImage = ImageIO.read(workOrderFile);
+            bufferedImage = ImageIO.read(workOrderFile);
             //直接处理二进制数据,避免因为文件扩展名错误导致的处理失败
-            //text = tesseract.doOCR(bufferedImage);
+            text = rapidOCR.run(bufferedImage).getStrRes().trim();
             log.debug("OCR解析{}到{}",workOrderFile.getAbsolutePath(),text);
         } catch (Exception e) {
             log.warn("OCR解析{}失败,case:{}",workOrderFile.getAbsolutePath(),e.getMessage());