boggerpiya

Text Detection & Extraction Project:

In this python project, we’re going to make a text detector and extractor from an image using opencv and ocr. We’ll use the Tesseract engine to perform the character recognition system and the pytesseract python package to interact with Tesseract in python.

What is OCR?

OCR or Optical Character Recognition is a system that can detect characters or text from a 2d image. The image could contain machine-printed or handwritten text. OCR can detect several languages, for example, English, Hindi, German, etc.

OCR is a widely used technology. Some popular real-world examples are:

Automatic number plate recognition.
At airports, passport recognition and information extraction.
Converting handwriting in real-time to control a computer (pen computing)

What is OpenCV?

OpenCV is an open-source computer vision library written in C/C++. It is mainly focused on image processing. OpenCV provides more than 2500 optimized algorithms. These algorithms can be used to detect and recognize faces & text, identify objects, track moving objects, etc.

What is Tesseract?

Tesseract is an optical image recognition engine that runs on various operating systems. It can detect more than 100 languages from all over the world. Tesseract is originally written in C/C++. But we are going to use it in python.

What is Pytesseract?

Python-tesseract is a wrapper for Tesseract-OCR Engine. It allows us to interact with the tesseract engine using python.

Challenges with Tesseract

Tesseract doesn’t perform well if the image contains a lot of noise.
If the font of any language is not trained then tesseract cannot detect that language accurately.
Image brightness or skewness may affect Tesseract’s performance as well.

Text Detection & Extraction Project Prerequisites:

Python – 3.x (we used Python 3.7.10 in this project)

OpenCV – v4.5

Please run below command to install the latest version of OpenCV:

pip install opencv-python

Tesseract Engine – v5.0.0

Please download the Tesseract engine executable (.exe) file and install it in “C:\Program Files\Tesseract-OCR\” directory.

Pytesseract – v0.3.7

Please run below command to install the latest version of Pytesseract:

Steps to Develop Text Detection & Extraction OpenCV Project :

Import necessary packages and configure Pytesseract with the Tesseract engine.
Reading images using OpenCV.
Detect text and numbers from the image.
Detect each character individually.
Create Bounding boxes over each character in the image.
Create a Bounding box over each detected text in the image.
Detect digits and alphabets individually.

Step 1 – Import necessary packages and configure Pytesseract with the Tesseract engine :

To work with the Tesseract engine first we need to configure pytesseract with the Tesseract Engine install path.

Explanation:

Configure Tesseract path with pytesseract.tesseract_cmd to run tesseract CLI (command line Interface) command via python and pytesseract functions.
If you installed Tesseract in a different location then make sure that you’ve entered the correct path.

Step 2 – Reading images using OpenCV :

We’ll use OpenCV for basic image operation

Explanation:

cv2.imread() function is for reading the image.
Tesseract engine works with RGB image and opencv reads the image as BGR . so we used cv2.cvtColor(img, cv2.COLOR_BGR2RGB) to convert the image BGR to RGB.

Step 3 – Detect & Extract Texts and Numbers from the Image:

We have the RGB image. So now we’ll detect & recognize texts and numbers from the image.

Explanation:

pytesseract.image_to_string() returns unmodified output as string from Tesseract OCR processing.

Output:

projectgurukul.org

allis better with friends
Saturday at 4pm

“A good friend is like a four-leaf
clover; hard to find and lucky to have.”

Step 4 – Detect and extract each character individually:

Now it’s time to detect and recognize each character individually. To do that we’ll use the image_to_boxes function. Let’s see how.

Explanation:

pytesseract.image_to_boxes returns recognized characters and their box boundaries.

In the output, we can see that the first column contains recognized characters, and the 2nd, 3rd, 4th, and 5th columns contain their boundaries.

Step 5 – Create Bounding boxes over each character in the image:

Now we have all the necessary information, so let’s create a bounding box and label each character in the image.

Explanation:

“shape” function returns a tuple containing information about image height, width and no. of channels.
Image_to_boxes return information as a string in each line. That’s why boxes.splitlines() function is used to separate each line and read each line individually.
Each line’s information is space-separated so split(” “) function is used to separate all information and then store each value individually.
Using OpenCv’s cv2.rectangle() function, create a bounding box over each character.
And then using cv2.putText() function labels each character in the image.

Step 6 – Create a Bounding box over each detected text in the image:

To create a bounding box over each text we need to detect each text individually as well as need bound coordinates for each text. To do this we are going to use the image_to_data function. First, see what the output data looks like.

Explanation:

image_to_data returns results containing boundaries, texts, and other important data.
We don’t need the first row that is the column name, that’s why count > 0 is used for not including the first row in the data processing.
split function splits each line and returns a list .
len(data)==12 is used because if you see the data, 12th column information is only available if any text is detected.
cv2.puttext and cv2.rectangle for labeling and creating bounding boxes around the text which we detect and extract.

Step 7 – Detect digits and texts individually:

We can detect and extract alphabets and digits individually, for example, if we need only alphabets then we can extract only alphabets, not numeric digits.

Explanation:

string.ascii_letters returns all the uppercase and lowercase alphabets in a single string.
Tessedit_char_whitelist is an argument of Tesseract which defines what characters the tesseract will detect.
To learn more about Tesseract visit: Tesseract documentation

So here we can see that only digits are detected, but there are some false results too.

Source Code:

Input:

Output:

Summary :

In this python project, we’ve built a text detector and extractor. Here we have used the Tesseract engine and pytesseract to perform all the text detection and extraction processes. We learned to work with the Tesseract engine and pytesseract function. Using these techniques we can make many cool projects related to OCR.

1.operating system:

2.What is operating system??

The system software that manage computer hardware ,software resources ,and provides common services for computer program. operating system is system software .An operating system is a program that, after being initially loaded into the computer by a boot program , manage all of the other application program in a computer.

1.Batch processing operating system: A job which has predefined sequence of commands, program and data as a single unit. Users had no direct interface with the system , and job execution was done in a batch system and users have to prepare a job separately to execute it. The operating system keeps a number a jobs in memory and executes them without any manual information , jobs are processed in order of submission. First come first served fashion..

2.Multitasking or Time sharing operating system:

In time sharing operating system allows the user to performs more than one task at a time, each task getting the same amount of time to execute. The operating system performs time sharing through CPU scheduling and multiprogramming. It is the division of CPU time for each process when more than one task are given by the user..

3.Distributed operating system:

A distributed operating system is system software , over a collection of independent, networked, communicating ,and physically separate computational nodes. Distributed system use many central processors to serve multiple real -time applications and users. As a result, data processing jobs are distributed between the processors..

4.Networking operating system:

A networking operating system is a computer operating system that is designed primarily to support workstations, personal computer and in some instances, older terminals that are connected on a local area network. The role of the NOS is to than provide basic network services and features that support multiple input requests simultaneously in multiuser environment..

5.Real time operating system: A real time operating system for real-time applications that processes data and events that have critically defined time constraints. A real time operating system time interval to process and respond to inputs is very small. Event driven systems switch between tasks based on their priorities, while time sharing systems switch the task based on clock interrupt ..

3.Structure of operating system

An operating system is a construct that allows the user application programs to interact with the system hardware.

1.simple structure:

These started as small system and rapidly expanded much further than their scope . A common example of this is MS-DOS.it was designed simply for a niche amount for people .An image to illustrate the structure of MS-DOS is as follows.

An image to illustrate the structure of MS-DOS is as follows.

MS-DOS

2.Layered structure of operating system:

In this ,the bottom layer is the hardware and the topmost layer is the user interface.

Each upper layer is built on the bottom layer. All the layers hide some structure, operations etc. from their upper layers .One problem with the layered structure is that each layer needs to be carefully defined. This is necessary because the upper layers can only use the functionalities of the layers below them.

4. Example of operating system: