java读取minst数据集保存到本地磁盘(已实现),存入磁盘速度过慢,该如何优化?具体代码如下

Question

package com.microservice.zjh.tushare.ionio.data;

import java.awt.Color;
import java.awt.image.BufferedImage;
import java.io.BufferedInputStream;
import java.io.File;
import java.io.FileInputStream;
import java.util.ArrayList;
import java.util.List;

import javax.imageio.ImageIO;

/**

• chenjh

*/
public class ReadMinst {

public static void main(String[] args) {
    // 1.读取文件
    File file = new File("src/main/java/com/microservice/zjh/tushare/ionio/data/minst/t10k-images.idx3-ubyte");

    try
    (
        //2.将文件注入文件输入流
        FileInputStream fis = new FileInputStream(file);
        //3.输入流注入到缓冲区
            //1)这是优点减少对磁盘的频率
        BufferedInputStream bis = new BufferedInputStream(fis);

    ) 
    {
        //4.设置缓冲区的大小
            //1)不需要在括号中写,小括号是写流的对象
        byte[] bytes = new byte[4];
        //5.读取到bytes中.从0,4 是输入 ,0,1,2,3
        bis.read(bytes, 0, 4);
        //6.bytes转hex[]
        //7.每个hex拼接Stringbuffer
        StringBuffer sb = new StringBuffer();
        for (int i = 0; i < bytes.length; i++) {
            byte b = bytes[i];
            //8.将转为无符号byte是int型,因为无符号byte区间是(0-255),所以在java中byte只能向上提升至int
            int unsignedbyte = b & 0xFF;
            //9.无符号byte转为String的hex(int转hex)
            String strHex = Integer.toHexString(unsignedbyte);
            //10.如果strHex长度为1,补0
            //11.拼接到StringBuffer,因为前四个值是32位
            if (strHex.length()<2) {
                sb.append(0);
            }
            sb.append(strHex);
        }
        System.out.println(sb);
        //11.StringBuffer转int,通过16进制字符串转为int
        int intfull = Integer.parseInt(sb.toString(),16);
        System.out.println(intfull);

        byte[] number_of_items = new byte[4];
        bis.read(number_of_items);
        StringBuffer number_of_items_sb = new StringBuffer();
        for (int i = 0; i < number_of_items.length; i++) {
            byte b = number_of_items[i];
            int unsignedbyte = b & 0xFF;
            String strHex = Integer.toHexString(unsignedbyte);
            if (strHex.length()<2) {
                number_of_items_sb.append(0);
            }
            number_of_items_sb.append(strHex);
        }
        System.out.println(number_of_items_sb);
        int number_of_items_int = Integer.parseInt(number_of_items_sb.toString(),16);
        System.out.println(number_of_items_int);
        bis.read(bytes, 0, 4);
        int xPixel = Integer.parseInt(bytesToHex(bytes), 16);           // 读取每行所含像素点数
        System.out.println(xPixel);
        bis.read(bytes, 0, 4);
        int yPixel = Integer.parseInt(bytesToHex(bytes), 16);   
        System.out.println(yPixel);
        //图片缓冲
        BufferedImage bi = new BufferedImage(28, 28, BufferedImage.TYPE_INT_BGR);
        for (int i = 0; i < number_of_items_int; i++) {
            // 1.保存图片到此地
            File saveinmnsit = new File("E:/mnsit/"+i+".png");
            if(saveinmnsit.exists()) {
                saveinmnsit.delete();
                saveinmnsit.createNewFile();
            }
            for (int y = 0; y < yPixel; y++) {
                for (int x = 0; x < xPixel; x++) {
                    int element = bis.read();
                    int rgb = new Color(element,element,element).getRGB(); 
                    if (element>127) {
                        System.out.print(1);
                    } else if(element<=127){
                        System.out.print(0);
                    }
                    bi.setRGB(x, y, rgb);
                }
                System.out.println();
            }
            //28*28跑完保存
            ImageIO.write(bi, "png", saveinmnsit);
            System.out.println();
            System.out.println();
            //执行几次,停止最外层运行for
            // if (i==0) {
            //     break;
            // }
        }
    } catch (Exception e) {
        e.printStackTrace();
    } 
}

public static String bytesToHex(byte[] number_of_items) {
    StringBuffer number_of_items_sb = new StringBuffer();
    for (int i = 0; i < number_of_items.length; i++) {
        byte b = number_of_items[i];
        int unsignedbyte = b & 0xFF;
        String strHex = Integer.toHexString(unsignedbyte);
        if (strHex.length()<2) {
            number_of_items_sb.append(0);
        }
        number_of_items_sb.append(strHex);
    }
    return number_of_items_sb.toString();
}

//黑底白字的图片

}
我想继续优化存入磁盘的操作,请问需要补充哪方面的知识,或者上面的代码该如何优化,请提供一下思路