Giter VIP home page Giter VIP logo

davegeneral / semi-auto-segmentation-labelling-system Goto Github PK

View Code? Open in Web Editor NEW

This project forked from a8039974/semi-auto-segmentation-labelling-system

0.0 1.0 0.0 22.35 MB

Author: Xu Liu Date: 02/18/2019 This is a semi-auto labelling software for those who work on image labelling. Input: Click the "Open" button, open the "image" directory and select the first image to start. After this step, the left canvas will display the original image with 50% transparence colored label on it. And the right canvas will display the labelled image with totally black ground and colorful foreground (one color, one instance). All the images in the "image" directory are the outputs(aka. prediction) of the segmentation deep neural network and are located at the "masks" directory. Most parts of the right image are labelled correctly by the neural network, what we need to do is just to revise it slightly. Instructions: 1. Click your right mouse button on the right image to pick a color (the color is corresponding to the pixel where your cursor locates) that you wish to revise on the left image. If there are instances that have not been labelled, you can click your right mouse button on the circular color palatte to pick a different color. 2. Move your cursor to right image, click at the place where you wish to revise the label, keep press the left mouse button and move it can draw curve lines, which can revise a large part. 3. The silder on the top of the two images can control the pen (or brush) size, when you need to paint (revise) a large part, you can move the slider a bit right to get a larger pen size. On the other hand, just move it left can be helpful for revising small part. 4. Click "Save" button on the right can save the revised image (the right one) as a 3 channels, 8 unsigned bits PNG format file at the "output" directory, which is in the same directory as the "images" directory. 5. Click "Next" button then the two canvas can refresh and load the next pair: image and label. Then just repeat the above operations.

QMake 2.10% C++ 47.06% Makefile 37.76% C 13.08%

semi-auto-segmentation-labelling-system's Introduction

Semi-Auto-Segmentation-Labelling-System

Author: Xu Liu
Date: 02/18/2019

This is a semi-auto labelling software for those who work on image labelling.

Input: Click the "Open" button, open the "image" directory and select the first image to start. After this step, the left canvas will display the original image with 50% transparence colored label on it. And the right canvas will display the labelled image with totally black ground and colorful foreground (one color, one instance).
All the images in the "image" directory are the outputs(aka. prediction) of the segmentation deep neural network and are located at the "masks" directory. Most parts of the right image are labelled correctly by the neural network, what we need to do is just to revise it slightly.

Instructions:

  1. Click your right mouse button on the right image to pick a color (the color is corresponding to the pixel where your cursor locates) that you wish to revise on the left image. If there are instances that have not been labelled, you can click your right mouse button on the circular color palatte to pick a different color.

  2. Move your cursor to right image, click at the place where you wish to revise the label, keep press the left mouse button and move it can draw curve lines, which can revise a large part.

  3. The silder on the top of the two images can control the pen (or brush) size, when you need to paint (revise) a large part, you can move the slider a bit right to get a larger pen size. On the other hand, just move it left can be helpful for revising small part.

  4. Click "Save" button on the right can save the revised image (the right one) as a 3 channels, 8 unsigned bits PNG format file at the "output" directory, which is in the same directory as the "images" directory.

  5. Click "Next" button then the two canvas can refresh and load the next pair: image and label. Then just repeat the above operations.

半自动分割标注系统说明手册 Semi-Auto Segmentation Labelling System Manual

整体布局: pic

初始界面: pic

软件说明书: 点击Instruction,显示如下: pic

选取图片: 仅需在images文件夹,即初始文件夹下选择图像(为方便批处理,最好选取第一张,主界面有next下一张快捷按钮),labels文件夹和images一一对应,图像会自动加载到主界面右边的canvas上; pic

和images文件夹对应的labels文件夹: pic

显示结果如下: pic

取色标注: 鼠标右键单击在右图中(或右上方调色板增加instance)选取颜色,再按住鼠标左键,在左图中移动可以重新修补标注(右图中会同步显示); 红色箭头:右击鼠标选取颜色; 紫色箭头:在左图中按住鼠标左键并移动可以重新标注,同时右边图像同步更新; pic

滑动条调节笔刷大小: pic

保存图像: pic pic

点击next进入下一张: pic

最后一张: pic

点击Exit退出程序: pic

semi-auto-segmentation-labelling-system's People

Contributors

earthonliner avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.