Scribd
Upload
39 views25 pages

Department of Electronics and Communication Engineering

The document summarizes a student project on comparing different edge detection algorithms for text extraction from images. The project used Prewitt, Sobel, and Robert's edge detection algorithms on images with varying text sizes, fonts, and languages. Prewitt edge detection provided the highest precision and recall rates. The project involved preprocessing images, applying edge detection and morphological operations, and extracting text characters. The techniques have applications in mobile banking, assisting visually impaired individuals, and other domains.

Uploaded by

Keerthy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
39 views25 pages

Department of Electronics and Communication Engineering

The document summarizes a student project on comparing different edge detection algorithms for text extraction from images. The project used Prewitt, Sobel, and Robert's edge detection algorithms on images with varying text sizes, fonts, and languages. Prewitt edge detection provided the highest precision and recall rates. The project involved preprocessing images, applying edge detection and morphological operations, and extracting text characters. The techniques have applications in mobile banking, assisting visually impaired individuals, and other domains.

Uploaded by

Keerthy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 25

SHRIDEVI INSTITUTE OF ENGINEERING AND

TECHNOLOGY
Sira road,TUMKUR-572106

Department of Electronics and Communication


Engineering
TEXT EXTRACTION FROM IMAGES USING DIFFERENT
EDGE DETECTION ALGORITHMS UNDER PERSPECTIVE OF
PRECISION AND RECALL RATE

Project Associate: Under the guidance of


Varuna Mrs. HARIPRIYA.R .BE,MTech
Manohar B S Asst.Professor, Dept of ECE
Keshav L S.I.E.T Tumkur-572106
Preetham S
 Digital images captured by camera contains many useful
information. Many research has been done for content
retrieval from natural images, for example, object
detection, face recognition and automatic text extraction.
 Three popular algorithms for this are: Edge detection,
Connected Component and texture-based. Prewitt Edge
Detector algorithms is implemented and applied to an
image with different text size, font styles and text language.
 the display boards though they are unfamiliar with the local
language Recent studies in the field of research a wide
variety of applications that require automated systems for
text extraction.
 One such recently developed application is the mobile
banking application provided by the banking institutions
that facilitates the customers to carry out the transactions
even on passing the image of the cheque to the server.
 All other such applications include tourist guide which
facilitate the tourists to people can understand that place
and Image text translation systems to help the visually
impaired people also tourists.
 “A Matlab Project in Optical Character Recognition (OCR)”, S Antani,
D.Crandall.
The goal of Optical Character Recognition (OCR) is to classify optical
patterns (often contained in a digital image) corresponding to other characters.
The process of OCR involves several steps including segmentation, feature
extraction, and classification.
 “Edge detection techniques”,H.K.Kim and Anil K.Jain
Detection determine which edge pixels should be discarded as noise and
which should be retained. Natural scene images contain text information which
is often required to be automatically recognized and processed.
 “A Survey On Various Approaches Of Text Extraction In Images”
Vijaykumar.K.L
Text Extraction from image is concerned with extracting the relevant text data
from collection of images. Rapid development of digital technology has resulted
in digitization of all categories of materials
 Many algorithms have been proposed for recognizing text data
in an image but results may be varying due to image size,
orientation, contrast and color.
 In non-document images, detecting text is more challenging
because of variation in text size and alignment. Most of
researchers have shown that edge detection method is not
successfully satisfied for text detection from natural scene
images.
 For people who are visually impaired the text to speech
conversion helps to know the information .
 And also translate the required language, from local language
to general language.
 For uneducated people also this text to voice conversion helps.
 Convert the image into monochrome image by thresholding.
 Filter the image for removing noise. Use Gaussian low-pass
filter.
 Apply edge-detector to the filtered image.
 Apply proper morphological operations, i.e. dilation to make
clusters of text regions.
 Multiply the resultant image with input black and white image
to get text in contrast with Plain background
Convolution mask Gx Convolution mask Gy

The gradient magnitude of prewitt operator is given by

| G |= ((Gx2+Gy2))1/2 or | Gx | + | Gy |
Convolution mask Gx Convolution mask Gy

The gradient magnitude of sobel operators is given by

| G |= ((Gx2+Gy2))1/2 or | Gx | + | Gy |
Convolution mask Gx Convolution mask Gy

The gradient magnitude of robert operator is given by

| G |= ((Gx2+Gy2))1/2 or | Gx | + | Gy |
 Pre-processing
 Gray-scaling
 Binarization
 Filtering
 Edge-detection
 Morphological Operation
 Character extraction
 Pre-processing
 Gray-scaling
 Binarization
 Filtering
 Edge-detection
 Morphological Operation
 Character extraction
 Mobile Banking
 Visually Impaired Person

 Vehicle License Plate extraction

 Text to voice

 Portable computers

 Text in WWW images

 Industrial automation
 Recognition of content in Natural Scene Images is trying for
complex foundation. There are numerous strategies accessible to
perform the content identification and character extraction in
regular scene pictures. Here the results obtained by Prewitt,sobel
and robert edge recognizable proof figuring on different pictures
Remembering the final objective to incorporate with that Prewitt
edge-pointer computation is healthier since it gives higher audit
rate as contemplated Sobel, Robert and altered Prewitt edge-
locator estimation.
 Giri, P. S.[2013], "Text Information Extraction And analysis from Image
Using Digital Image Processing Techniques"
 Agarwal P, Varma R, “Text Extraction From Images”, IJCSET
 M. Swamy Das, B. HimaBindhu, A. Govardhan, “Evaluation of Text
Detection and Localization Methods in Natural Images.”
 Jung K, Kim KI, Jain AK,“Text Information Extraction in Images and
Video”
 Dutoit, T., 1993. High quality text-to-speech synthesis of the French
language. Doctoral dissertation, Faculte Polytechnique de Mons.
 Suendermann, D., Höge, H., and Black, A., 2010. Challenges in Speech
Synthesis. Chen, F., Jokinen, K., (eds.), Speech Technology, Springer
Science + Business Media LLC.
 Allen, J., Hunnicutt, M. S., Klatt D., 1987. From Text to Speech: The
MITalk system. Cambridge University Press.
THANK YOU

You might also like