Update README.md

README.md

@@ -28,15 +28,12 @@ These is example of the output produced from the current code.
 # Dependencies
 * OpenCV: 3.0 or newer.
 
-# Build
+# Usage as an exe
 ```Console
 mkdir build
 cd build
 cmake ..
 make
-```
-# Usage
-```Console
 ./Image2Lines the_path_to_the_img
 ```
 the output will be:
@@ -45,11 +42,34 @@ the output will be:
 * 'Final_lines.jpg' : representing the final lines drawn on the original image.
 * 'Contours.jpg' : represnting the detected contours.
 
-# TODO
-* Improve detecting peaks and valleys function.
-* Use a better binarization technique than the currrently used. (I will use my other repository named Binarization".
+# Usage in your code
+You should add LineSegmentation.cpp and LineSegmentation.hpp to your project.
+```C++
+    #include "LineSegmentation.hpp"
+
+    // Create an object from LineSegmentation and give to it the image path
+    LineSegmentation line_segmentation(img_path);
+
+    // Run segment() method to segment lines. The output will be a vector of CV::Mat each representing a line.
+    vector<cv::Mat> lines = line_segmentation.segment();
+
+```
+**NOTE** You should link correctly OpenCV in your working project. if you use CMake see CMakeLists.txt file to see how to add opencv.
 
-## Database used
-IAM database.
+# Development
+# Build 
+```Console
+mkdir build
+cd build
+cmake ..
+make
+```
+# Current issues
+* Finding peaks and valleys can cause in some cases the lines to be falsely segmented. If you can improve that problem, feel free to make a pull request.
 
+# TODOs
+* Improve detecting peaks and valleys function.
+* Use a better binarization technique than the currrently used (OTSU) in order to get better binarization in case of camera images. (I will use my other repository named Binarization).
 
+## Database used
+IAM database (Educational use permit).