Vi Tools App

This app allows user to:

• Load dataset containing jpgs and their xmls,
• Create dataset mirrored copy,
  • 
• Translate xml files to txt files which are understanded by darknet

Workflow

1. Use labelImg software to create labels for dataset
2. Use Vi Tools App -> Mirror IMGs function to create mirror of your dataset
3. Use Vi Tools App -> Create TXT files function to translate xmls to txt for darknet

Links you need to know!

• https://github.com/tzutalin/labelImg

• https://github.com/AlexeyAB/darknet

Technologies

App supports:

• Asynchronus programming
• Colorful status presentation
• Mvvm

Xml file structure

<annotation>
    <folder>folder</folder>
    <filename>file name</filename>
    <path>jpg path</path>
    <source>
        <database>Unknown</database>
    </source>
    <size>
        <width>864</width>
        <height>1159</height>
        <depth>1</depth>
    </size>
    <segmented>0</segmented>
    <object>
        <name>class name</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>intiger</xmin>
            <ymin>intiger</ymin>
            <xmax>intiger</xmax>
            <ymax>intiger</ymax>
        </bndbox>
    </object>
</annotation>

Txt file structure

[class numer] [Left] [Top] [Width] [Height]

Class numbers are intigers, first one has number '0'. All values are independent from resolution and are given in percents. 0.52502 means that obiect is located in 52,502% of images height.

1 0.52502 0.46065 0.09259 0.68766

Worth mentioning that next labels will apear in next rows. Their amount is corresponding to 'object' elements in xml file. Like so:

1 0.52502 0.46065 0.09259 0.68766
1 0.52502 0.46065 0.09259 0.68766
1 0.52502 0.46065 0.09259 0.68766

And so on.

App overview