Segmentation in OCR !!

A basic explanation of different levels of Segmentation used by the OCR system. Welcome to part-III about the working of the OCR system.

Welcome to part-III about the working of the OCR system. In the previous post, we have seen the basic and widely used Preprocessing techniques in the OCR system.

In this part, we are going to cover the techniques of another most important phase of the OCR system, Segmentation.

Segmentation is nothing but breaking the whole image into subparts, to further process them.

Segmentation of image is done in the following sequence :

→ Line level Segmentation

→ Word level Segmentation

→ Character level Segmentation

For all the above-mentioned levels of Segmentation, we are going to use the Histogram Projection technique.

Histogram Projection Method

Let me give you a brief introduction to the Histogram Projection method. Once the coloured image is converted to the binary image only black and white pixels are present in the image.

In a binary image, pixels which represent the useful information are called _Foreground pixels, _whereas the pixels which are not foreground pixels are called Background pixels. It is our choice whether a foreground pixel should be a white pixel or black pixel while binarizing the image.

_Horizontal Histogram Projection: _In this method, we count the No.of foreground pixels along the rows of the image and the resultant array is of the size equal to No.of rows in the image (Height of the image).

programming ocr image-processing machine-learning segmentation deep learning

Changing Image Backgrounds Using Image Segmentation. Make the world your green screen. Hello readers! In this article I’ll be discussing another cool trick we can do with deep learning.

Being a practitioner in Machine Learning, you must have gone through an image classification, where the goal is to assign a label or a class to the input image.

Implementing an image segmentation model that changes backgrounds. Hello seekers! In this post (part 2 of our short series — you can find part 1 here), I’ll explain how to implement an image segmentation model with code. This model will allow us to change the background of any image, just by using the API that we’ll build.

During my studies at JKU there was a task for preprocessing images for a machine learning project. It is necessary to clean the raw images…

We supply you with world class machine learning experts / ML Developers with years of domain experience who can add more value to your business.