From 56ccb39539d17ceedbfc6e69d6816971ac3c57e8 Mon Sep 17 00:00:00 2001
From: Clemens Neudecker <952378+cneud@users.noreply.github.com>
Date: Tue, 11 Oct 2022 11:37:00 +0200
Subject: [PATCH] Update README

* recommend cropping (fix #49)
* document huggingface saved_model
---
 README.md | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/README.md b/README.md
index d5cff03..d1f5c8e 100644
--- a/README.md
+++ b/README.md
@@ -8,7 +8,7 @@
 
 ## Introduction
 
-This tool performs document image binarization using trained models. The method is based on [Calvo-Zaragoza and Gallego, 2018](https://arxiv.org/abs/1706.10241).
+This tool performs document image binarization using a trained ResNet50-UNet model. 
 
 ## Installation
 
@@ -18,10 +18,14 @@ Clone the repository, enter it and run
 
 ### Models
 
-Pre-trained models can be downloaded from here:   
+Pre-trained models in  `h5` format can be downloaded from here:   
 
 https://qurator-data.de/sbb_binarization/
 
+We also provide a Tensorflow `saved_model` via Huggingface:
+
+https://huggingface.co/SBB/sbb_binarization
+
 ## Usage
 
 ```sh
@@ -32,9 +36,11 @@ sbb_binarize \
   <output image>
 ```
 
-**Note** In virtually all cases, applying the `--patches` flag will improve the quality of results.
+In virtually all cases, applying the `--patches` flag will improve the quality of results.
+
+Images containing a lot of border noise (black pixels) should be cropped beforehand to improve the quality of results.
 
-Example
+### Example
 
 ```sh
 sbb_binarize --patches -m /path/to/models/ myimage.tif myimage-bin.tif