webshotcompare
v1.3.1
Published
Image comparison and offset detection optimized for web screenshots.
Downloads
15
Maintainers
Readme
💡 Quick overview
Image comparison and offset detection optimized for web screenshots. Why is this useful?
- 输出可视化差异/偏移图片结果(visualize difference/offset image results)
- 可调整的阈值(adjustable thresholds)
- 优化的OpenCV.js(volume-optimized OpenCV.js)
- 专为页面测试优化的差异算法(diff algorithm optimized for screenshot)
- 稀少的OpenCV.js实践参考代码(easy-to-read OpenCV.js practice code)
| SAMPLE | RESULT.png | RESULT_MATCHES.png |RESULT_DIFF.png | | :-----:| :----: | :----: |:----: | | 1 | | | | | 2 | | | |
sample 1 output:
{
diffPass: false,
shiftPass: false,
diff: { diffPercent: 5.39483400081052, imgDistance: 0.03125 },
shift: { rate: 0.5106382978723404, shiftCount: 24, goodMatchesCount: 47 },
img: {
diff: '/Users/key/Desktop/mini/testForImgCom/images/RESULT.png',
shift: '/Users/key/Desktop/mini/testForImgCom/images/RESULT_MATCHES.png',
diffByPixel: '/Users/key/Desktop/mini/testForImgCom/images/RESULT_DIFF.png'
}
}
sample 2 output:
{
diffPass: true,
shiftPass: true,
diff: { diffPercent: 0.49524, imgDistance: 0 },
shift: { rate: 0, shiftCount: 0, goodMatchesCount: 464 },
img: {
diff: '/Users/key/Desktop/mini/testForImgCom/images/RESULT.png',
shift: '/Users/key/Desktop/mini/testForImgCom/images/RESULT_MATCHES.png',
diffByPixel: '/Users/key/Desktop/mini/testForImgCom/images/RESULT_DIFF.png'
}
}
give a star: PapriKey/webshotcompare
🎁 Usage:
const webshotCompare = require("webshotcompare");
async function run() {
let res = await webshotCompare(
"/Users/key/Desktop/WechatIMG1634.png",// 原图
"/Users/key/Desktop/WechatIMG1635.png",// 需要对比的图
"/Users/key/Desktop/mini/js_dip_solve/images", //保存路径
{
debug: false,//此处添加需要修改的config
}
);
console.log(res);
}
run();
🪄 Output/return:
{
diffPass: false,//是否通过差异检测
shiftPass: false,//是否通过偏移检测
diff: { diffPercent: 5.39483400081052, imgDistance: 0.03125 },//差异检测量(像素差异百分比,欧氏距离值)
shift: { rate: 0.5106382978723404, shiftCount: 24, goodMatchesCount: 47 },//偏移检测结果(偏移像素点/匹配点(总样本)比例,偏移点数量,总样本数量)
img: {
diff: '/Users/key/Desktop/mini/js_dip_solve/images/RESULT.png',
shift: '/Users/key/Desktop/mini/js_dip_solve/images/RESULT_MATCHES.png',
diffByPixel: '/Users/key/Desktop/mini/js_dip_solve/images/RESULT_DIFF.png'
}
}
🔧 Config:
{
//pixelMatch即diff的默认配置
pixelMatch_threshold: 0.05, //pixelMatch的阈值
pixelMatch_diffColor: [255, 255, 255], //pixelMatch的颜色
pixelMatch_diffMask: [0, 0, 0], //pixelMatch的遮罩层颜色
pixelMatch_DiffPercentThreshold: 1, //像素级对比差异的阈值(1-100),超过则diff不通过<---------1
pixelMatch_imgDistanceThreshold: 0.01, //图片欧氏距离阈值,超过则diff不通过<------------------2
binThreshold_threshold: 100, //二值化的像素阈值
morphologyEx_kernelSize: 20, //形态学操作的核大小
rectangleColor: [255, 0, 0, 255], //图像差异矩形的颜色
knnDistance_threshold: 0.3, //knn的欧式距离阈值(判断两幅图片中两匹配点是否匹配)
shiftJudge_shiftThreshold: 5, //两匹配点间像素x/y坐标偏移阈值
shiftJudge_rateThreshold: 0.3, //偏移点与总样本间比例阈值,超过则shift不通过<-------------------3
debug: false, //是否输出调试信息和中间图片
timeOut : 10000 //超时时间(毫秒)
};