Survey on Deep Learning Method for Designing Text Based Captcha
MALAVIKA R
, MANJUSHA NAIR S
deep learning, convolutional neural network, VGGNet, Neural style transfer
Since captcha’s discovery, it is the most widely used security tool. They are of many types. Text based captchas are most commonly used. Large type of early easy captchas were easily infiltrated. After many number of modifications and changes ,existing types came into existence. There are different resistive techniques like Crowding Characters Together (CCT), Noise arcs, Complicated backgrounds ,Hollow schemes and Two layer structures, but all of these have drawbacks and can be broken. Here we are trying to design an effective text captcha that resist the existing attacks against captcha. Captcha is based on two principle, anti segmentation and anti recognition. Segmentation makes a captcha weak. The proposed method to design captcha is fast and effective with deep learning techniques.
"Survey on Deep Learning Method for Designing Text Based Captcha", IJSDR - International Journal of Scientific Development and Research (www.IJSDR.org), ISSN:2455-2631, Vol.4, Issue 3, page no.477 - 480, March-2019, Available :https://ijsdr.org/papers/IJSDR1903081.pdf
Volume 4
Issue 3,
March-2019
Pages : 477 - 480
Paper Reg. ID: IJSDR_190259
Published Paper Id: IJSDR1903081
Downloads: 000347186
Research Area: Engineering
Country: KOLLAM, Kerala, India
ISSN: 2455-2631 | IMPACT FACTOR: 9.15 Calculated By Google Scholar | ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.15 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJSDR(IJ Publication) Janvi Wave
