Hand Written Digit Recognition

idea of works

Handwritten recognition enables us to convert handwriting documents into digital forms. Set an image that is content a handwritten digit then detects it and converts to ASCII that can be known by the computer.

It's a hard task for a computer because there are many types of handwritten digits. The accuracy and matches of detects 90%-98%, can obtain this by test knowns of samples then check the correct average results.

Training part (classification):

First, we need a preset of handwritten number as images.

1. convert image to gray scale cv::cvtColor(IMG0, IMG0_GRAY,cv::COLOR_BGR2GRAY); ///cvt to GRAY color

2. to binary cv::adaptiveThreshold(IMG0_GRAY, IMG0_BIN,255, cv::ADAPTIVE_THRESH_GAUSSIAN_C, cv::THRESH_BINARY_INV, 11,2);

3. find contours (i.e number in image) cv::findContours(IMG0_BIN, Contours, optional, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_SIMPLE);

4. cut each contour detect in the image and display for select which is the number belongs to, after selecting a number from the keyboard

5. listNumberDetect.push_back(intChar); save pressed number as xml . [https://raw.githubusercontent.com/MohammadTeeU/hand-written-digits/master/FINAL-PROJ/xmlimgs/listNumberDetectF.xml]

6. convert contour to float row cv::Mat imgOneRow = matImageFloat.reshape(1, 1);

7. ImgsDetectd.push_back(imgOneRow); save data of contour (or number). as xml file [https://raw.githubusercontent.com/MohammadTeeU/hand-written-digits/master/FINAL-PROJ/xmlimgs/ImgsDetectdF.xml]

We trained about 500 samples of (0-9)

Regression we used a supervised learning approach (K-nearest neighbors KNN algorithm).

1. first read XMLs files from training FILE0["listNumberDetect"] >> NumberIsDetectd; FILE1["fsImgsDetectd"] >> RowOfImgsDetect;

2. train each number with each data got form xml files kNearest->train(RowOfImgsDetect,cv::ml::ROW_SAMPLE,NumberIsDetectd);

3. read image and convert to binary for detect number in it .

4. find contours in image cv::findContours(IMG0_BIN,Contours, OPT, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_SIMPLE);

5. convert contour to float row cv::Mat matRow = matFloat.reshape(1, 1);

6. get nearest of number with data-set we had kNearest->findNearest(matRow, 23, matAsChar);

7. matAsChar number that found

ui flow

---

Requirements:

• Qt5 (http://qt.io)
• OpenCV4 (https://opencv.org)
• compilers:
  1. for linux: gcc or clang
  2. for android: clang with (SDK & NDK)
  3. for windows: MSVC or mingw compiler

Flow of work

1-Draw Number on canvas:

---

2-Click on test to get result:

---

3-Select Image form local storage:

---

4-Get result:

---

5-Take picture from Camera:

---

6-Get result:

---