java opencv 4.0.1安装配置

如果没有把dll扔到jdk会报错

Exception in thread "AWT-EventQueue-0" java.lang.UnsatisfiedLinkError: no opencv_java401 in java.library.path
    at java.lang.ClassLoader.loadLibrary(ClassLoader.java:)
    at java.lang.Runtime.loadLibrary0(Runtime.java:)
    at java.lang.System.loadLibrary(System.java:)
    at com.hs.jframe.Jframe.jButton5ActionPerformed(Jframe.java:)
    at com.hs.jframe.Jframe.access$(Jframe.java:)
    at com.hs.jframe.Jframe$.actionPerformed(Jframe.java:)
    at javax.swing.AbstractButton.fireActionPerformed(AbstractButton.java:)
    at javax.swing.AbstractButton$Handler.actionPerformed(AbstractButton.java:)
    at javax.swing.DefaultButtonModel.fireActionPerformed(DefaultButtonModel.java:)
    at javax.swing.DefaultButtonModel.setPressed(DefaultButtonModel.java:)
    at javax.swing.plaf.basic.BasicButtonListener.mouseReleased(BasicButtonListener.java:)
    at java.awt.Component.processMouseEvent(Component.java:)
    at javax.swing.JComponent.processMouseEvent(JComponent.java:)
    at java.awt.Component.processEvent(Component.java:)
    at java.awt.Container.processEvent(Container.java:)
    at java.awt.Component.dispatchEventImpl(Component.java:)
    at java.awt.Container.dispatchEventImpl(Container.java:)
    at java.awt.Component.dispatchEvent(Component.java:)
    at java.awt.LightweightDispatcher.retargetMouseEvent(Container.java:)
    at java.awt.LightweightDispatcher.processMouseEvent(Container.java:)
    at java.awt.LightweightDispatcher.dispatchEvent(Container.java:)
    at java.awt.Container.dispatchEventImpl(Container.java:)
    at java.awt.Window.dispatchEventImpl(Window.java:)
    at java.awt.Component.dispatchEvent(Component.java:)
    at java.awt.EventQueue.dispatchEventImpl(EventQueue.java:)
    at java.awt.EventQueue.access$(EventQueue.java:)
    at java.awt.EventQueue$.run(EventQueue.java:)
    at java.awt.EventQueue$.run(EventQueue.java:)
    at java.security.AccessController.doPrivileged(Native Method)
    at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:)
    at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:)
    at java.awt.EventQueue$.run(EventQueue.java:)
    at java.awt.EventQueue$.run(EventQueue.java:)
    at java.security.AccessController.doPrivileged(Native Method)
    at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:)
    at java.awt.EventQueue.dispatchEvent(EventQueue.java:)
    at java.awt.EventDispatchThread.pumpOneEventForFilters(EventDispatchThread.java:)
    at java.awt.EventDispatchThread.pumpEventsForFilter(EventDispatchThread.java:)
    at java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:)
    at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:)
    at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:)
    at java.awt.EventDispatchThread.run(EventDispatchThread.java:)

首先引入依赖

    <!-- https://mvnrepository.com/artifact/org.bytedeco.javacpp-presets/opencv -->
<dependency>
    <groupId>org.bytedeco.javacpp-presets</groupId>
    <artifactId>opencv</artifactId>
    <version>4.0.-1.4.</version>
</dependency>

再去官网下载一些东东https://opencv.org/releases/

因为maven仓的版本 最高只有4.0.1所以，下载的时候，找到4.0.1下载

官网下载资源

然后运行，解压后会得到一些文件

然后把dll扔到jdk的bin文件夹下

这个路径具体看自己的实际情况

然后构建一下，看到有这个jar包就行了，

开始之前呢，需要先了解一下opencv的行列方向问题

一般图片都需要进行灰度化然后二值化，

参考下图

这样基本上也可以识别了，但是还是有一些噪点，最好还是去一下噪点

对于分辨率比较低的图片，噪点基本都是一两个像素

简单的办法就是计算黑色像素点周围的8个像素是不是全部白色，如果是那8个像素点的值加起来就是2040，就把这个点处理成白色，如果不是，那可能就不是噪点

效果还凑合

代码如下

  //因为去了边框，所以，行列数，做减2处理，即不处理边框上的像素
         for(int x=1;x<col-1;x++){
             for(int y=1;y<row-1;y++){
                 double[] binDataList=target.get(y, x);
                 double binData=binDataList[0];

                 if(binData==0){//黑色
                     double[] x1list=target.get(y-1, x-1);
                     double[] x2list=target.get(y-1, x);
                     double[] x3list=target.get(y-1, x+1);

                     double[] x4list=target.get(y, x-1);
                     double[] x6list=target.get(y, x+1);

                     double[] x7list=target.get(y+1, x-1);
                     double[] x8list=target.get(y+1, x);
                      double[] x9list=target.get(y+1, x+1);
                      double x1=x1list[0];
                      double x2=x2list[0];
                      double x3=x3list[0];
                      double x4=x4list[0];
                      double x6=x6list[0];
                      double x7=x7list[0];
                      double x8=x8list[0];
                      double x9=x9list[0];
                      if(x1+x2+x3+x4+x6+x7+x8+x9>=2040){
                          System.out.println("坐标xy："+x+","+y+"此点是噪点，抹除处理");
                          target.put(y, x, 255);//画一个白点
                      }
                 }
                 else{

                 }
             }

         }

稍微复杂一点的有连通域计算面积办法

https://blog.csdn.net/marooon/article/details/81332487