Text and non-text detection and text recognition in an image

(1 customer review)



In this project, we have successfully developed a MATLAB GUI with automatic book cover detection and recognition to help people take the advantages of both online shopping and real book store shopping, with emphasis on adding pre-process and post-process steps to improve the performance of OCR functions, and finally increase the rate of correct recognition.

The GUI we developed is platform-invariant, robust, fast and accurate in providing good user interaction experience. Although the initial motivation for this project is to provide better experience for readers, but this system can also be used to detect other text based items such as CD covers and newspapers. With good flexibility, the system can be further extended to be functional for multiple languages and implemented on mobile devices.



%% Step 1: Detect Candidate Text Regions Using MSER

%% Step 2:Remove Non-Text Regions Based On Basic Geometric Properties

%% Step 3: Remove Non-Text Regions Based On Stroke Width Variation

%% Step 4: Merge Text Regions For Final Detection Result

%% Get rid of the text where bounding boxes are not aligned together

%% Step 5: Recognize Detected Text Using OCR

Arduino code for detection of muscle fatigue and muscle growth

OpenCV code for text detection


1 review for Text and non-text detection and text recognition in an image

  1. blank


    I think your site will help a lot of students to do their projects better.

Add a review

Your email address will not be published. Required fields are marked *