LBChecker
LBChecker is a tool to extract value from YOLO format data text files and draw bounding boxes in clean images.
Purpose
- We want to detect objects from images and videos, we choose YOLO as the model
- To train this model, we will need a large of datasets
- Most of time, a dataset with good quality can take improve the models training results
- So, how can we ensure the quality of datasets and check whether we draw the bounding boxes correctly or not
- This tool can read the text files and convert the value to produce x_center y_center width height of the bounding boxes and draw it to the images
Before we start
To use this tool, make sure you already install the following:
- PHP >=7.0
- Fileinfo Extension
- GD Library (>=2.0)
Quick start
- Move the images and text files into the input folder
- Run main.php
- Check the result in output folder
Some tips to know
- To change or add more classes of the labels
$labelTitleMapping = [
0 => 'smoking',
1 => 'reading',
2 => 'calling',
3 => 'drinking',
4 => 'eating',
- To change the color of the bounding boxes
$labelColorMapping = [
0 => 'rgba(255, 157, 151, 0.5)',
1 => 'rgba(145, 205, 49, 0.5)',
2 => 'rgba(51, 67, 145, 0.5)',
3 => 'rgba(145, 205, 49, 0.5)',
4 => 'rgba(255, 48, 195, 0.5)',