pqpo/SmartCropper
An Android image cropping library that uses a TensorFlow HED neural network to detect document borders and correct perspective.

SmartCropper is an image cropping library for Android that automatically identifies borders in photos of documents, ID cards, and business cards. It employs a Holistically-Nested Edge Detection (HED) network implemented with TensorFlow to replace traditional Canny edge detection, improving border recognition accuracy. The library uses OpenCV for perspective transformation to crop and correct skewed documents, providing manual adjustment controls with magnification for fine-tuning selection areas.
Frequently asked
- What is pqpo/SmartCropper?
- An Android image cropping library that uses a TensorFlow HED neural network to detect document borders and correct perspective.
- Is SmartCropper open source?
- Yes — pqpo/SmartCropper is an open-source project tracked on heatdrop.
- What language is SmartCropper written in?
- pqpo/SmartCropper is primarily written in Java.
- How popular is SmartCropper?
- pqpo/SmartCropper has 4.1k stars on GitHub.
- Where can I find SmartCropper?
- pqpo/SmartCropper is on GitHub at https://github.com/pqpo/SmartCropper.