This application is developed for easier data preparation for YOLO. The application can do (after loading all images in folder) 1-Create Images Annotation, and create a rotated 90 degree version to increase the dataset. 2-Play Dataset with different speed with the ability to show only images with annotated objects or images without any annotation or both. 3-Change Class number for all images, and text file with all 90% of images for training and 10% of images for test. 4-show statistics of all images in the folder ( how many images, images with object count and images without object count) Future work 1-ability to create project with all images of different classes 2-create all needed yolo files: a-yolocfg file modifications b-names and data files. c-annotate from video.
vpattanaik / imageyolomanipulation Goto Github PK
View Code? Open in Web Editor NEWThis project forked from moutazaboseada/imageyolomanipulation
Images , annotation, load to check annotations, change class to be prepared for YOLO