Description

- RUN LENGTH SMOOTHING ALGORITHM(RLSA) is a method mainly used for block segmentation and text discrimination.
- It mainly uses in Document Image Processing to extract out the ROI(region of interest) like block-of-text/title/content with applied heuristics.

Input & Output

                                                            output of 3 cases with value "10" are in the below image

Sample Test Case

- value = 3
- input - [0, 0, 255, 255, 255, 0, 0, 255, 0, 0, 255, 0, 255]
- output - [0, 0, 255, 255, 255, 0, 0, 0, 0, 0, 0, 0, 255]

Unittest Results

Prerequisites

- python3.5+
- Image must be a binary ndarray(255's/1's/0's)
- Must pass a predefined limit, a certain integer "value"

How it works

- '255'(white pixel) wil be converted to '0'(black pixel) in a image, if the number of adjacent 255's are less than the predefined limit "value".
- The "value" varies among the different images.

Install requirements

- pip install -r requirements.txt
- if python.__version__ < 3.5: pip install typing
- Read the docs here !(https://pypi.org/project/typing/)

Method

- rlsa

Parameters

- image - numpy.ndarray(required)
- horizantal - boolean(required)
- vertial - boolean(required)
- value - any positive integer(int)(required)

IPython Usage to convert Image to Binary and RLSA application

- # convert the image to binary
- import cv2
- image = cv2.imread('image.jpg')
- gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
- (thresh, image_binary) = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)
- # function call
- import rlsa
- image_rlsa_horizontal = rlsa.rlsa(image_binary, True, False, 10)