Image Annotation Tools

• FIAT
• Sloth http://sloth.readthedocs.io/en/latest/first_steps.html
• LabelIMG

Fast Image Data Annotation Tool (FIAT)

FIAT enables image data annotation, data augmentation, data extraction, and result visualisation/validation.

• annotate images for image classification, optical character reading (digit classification, letter classification), ...
  
• extract data into different format (Caffe LMDB, OpenCV Cascade Classifiers, Tesseract ... ) with data augmentation (resizing, noise in translation / rotation / scaling, pepper noise , gaussian noise, rectangle merging, line extraction ...)
  

The philosophy of this tool is

• to enable fast annotation : annotate data just by selecting the diagonal of the object, for a fixed ratio. Press Enter if the class is always the same. Type the letter of the class otherwise,
  
• to be re-usable for different scenarios, and leave you free to build any other tool as input of the annotation process, using any pre-existing weaky classifier (depending on your case) or bounding box proposal algorithm such as selective search, to facilitate annotation with a list of rectangles to select or discard, by just typing the letter of the class or ESCAPE KEY,
  
• to have the extraction tool act like a monad, so that you can apply transformation in any order, at any stage of your process, the format remaining the same : given a directory and a CSV file, extraction tool will produce a new directory and CSV file, in the same format, by default,
  
• to feed any classification / training process,
  
• to be usable for visualisation and export : a visual check that the data is correctly annotated, after manual annotation, extraction, or even after your own bounding box prediction algorithm if it uses the Output class to its produce results.
  

Requires OPENCV 3 and Google Protobuf.
https://github.com/christopher5106/FastAnnotationTool

LabelImg is a graphical image annotation tool and label object bounding boxes in images 

https://github.com/tzutalin/labelImg