如何将base64图像压缩为自定义大小

how to compress a base64 image to custom size(如何将base64图像压缩为自定义大小)

问题描述

我使用 base64 发送/接收我的图像.我有一个 base64 字符串，我想将它压缩到我的大小.

I send/receive my image by using base64. I have a base64 string and I want to compress it to my size.

例如，我想将照片大小减小到 100kb.

for example I want to reduce photo size to 100kb.

有可能吗?

推荐答案

这是一个有趣的挑战，因为它涉及到二进制搜索，直到找到合适的大小.我不会建议你用 base64 而不是 blob 来解决这个问题，因为你应该将它作为二进制 (blob) 处理，否则它会占用大约 33% 的 base64 数据

This was a fun challenge cuz it involved a binary search until it finds the right size. I'm not going to advice you to solve this with base64 instead of blob cuz you should really handle it as binary (blob) otherwise it takes up ~33% more data as base64

此代码包括调整大小，您可以设置最大宽度/高度，并且仍然能够保持纵横比和自动质量查找，直到找到与 MAX_SIZE 匹配的正确质量

This code includes resizing that you can set a max width/hight and still be able to keep the aspect ratio and auto quality lookup until it finds the correct quality to match the MAX_SIZE

console.log('Downloading lorem ipsum image to simulate a file from user input')

fetch('https://picsum.photos/1920/1080/?random')
.then(res => res.blob())
.then(blob => {
  const img = new Image()
  img.src = URL.createObjectURL(blob)

  console.log(`Original image size (at 1920x1080) is: ${blob.size} bytes`)
  console.log('URL to original image:', img.src)
  
  img.onload = () => resize(img, 'jpeg').then(blob => {
    console.log('Final blob size', blob.size)
    console.log('Final blob url:', URL.createObjectURL(blob))

    console.log('
Now with webp
')

    resize(img, 'webp').then(blob => {
      console.log('Final blob size', blob.size)
      console.log('Final blob url:', URL.createObjectURL(blob))
    })
  })
}) 


const MAX_WIDTH = 1280
const MAX_HEIGHT = 720
const MAX_SIZE = 100000 // 100kb

async function resize(img, type = 'jpeg') {
  const canvas = document.createElement('canvas')
  const ctx = canvas.getContext('2d')
  
  ctx.drawImage(img, 0, 0)
  
  let width = img.width
  let height = img.height
  let start = 0
  let end = 1
  let last, accepted, blob
  
  // keep portration
  if (width > height) {
    if (width > MAX_WIDTH) {
      height *= MAX_WIDTH / width
      width = MAX_WIDTH
    }
  } else {
    if (height > MAX_HEIGHT) {
      width *= MAX_HEIGHT / height
      height = MAX_HEIGHT
    }
  }
  canvas.width = width
  canvas.height = height
  console.log('Scaling image down to max 1280x720 while keeping aspect ratio')
  ctx.drawImage(img, 0, 0, width, height)
  
  accepted = blob = await new Promise(rs => canvas.toBlob(rs, 'image/'+type, 1))
  
  if (blob.size < MAX_SIZE) {
    console.log('No quality change needed')
    return blob
  } else {
    console.log(`Image size after scaling ${blob.size} bytes`)
    console.log('Image sample after resizeing with losseless compression:', URL.createObjectURL(blob))
  }
  
  // Binary search for the right size
  while (true) {
    const mid = Math.round( ((start + end) / 2) * 100 ) / 100
    if (mid === last) break
    last = mid
    blob = await new Promise(rs => canvas.toBlob(rs, 'image/'+type, mid))
        console.log(`Quality set to ${mid} gave a Blob size of ${blob.size} bytes`)
    if (blob.size > MAX_SIZE) { end = mid }
    if (blob.size < MAX_SIZE) { start = mid; accepted = blob }
  }

  return accepted
}

PS/警告 如果您在画布元素上绘制 jpg 图片并在没有调整大小、操作质量损失或更改格式的情况下恢复图像，则 Canvas 不会进行任何良好的压缩 toBlob('image/jpg', cb, 1) 那么你肯定会得到一个更大的文件，因为它们可能已经被很好地压缩并且画布没有剂量.我只改变质量和使用画布 api 减小大小的最大宽度/高度.您需要一些压缩器来进一步减少它而不会造成质量损失.

PS/warning Canvas don't do any good compression, if you paint a jpg picture on a canvas element and get the image back with no resizing, manipulation quality loss or changing the format toBlob('image/jpg', cb, 1) then you will most definitely get a larger file back since they probably already are well compressed and canvas dose none. I only change the quality & max width/height to reduce the size with the canvas api. You would need some compressor to reduce it even more without quality loss.

• jsfiddle 用画布增加文件的演示
• imageoptim
• squoosh.app
• zopfli
• pngcrush

这篇关于如何将base64图像压缩为自定义大小的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持编程学习网！