Real-Time Cartoonization Using Raspberry Pi  
  Authors : Parimal Sikchi; Neha Beknalkar; Swapnil Rane

 

Cartoonization is a method of image stylization in which input moreover looks like a sketch or cartoon like image. It includes making edges bolder, making colors brighter and lively, accompanied with smoothening operation so as to filter out the high frequency details. Furthermore, reduction in the luminance contents is carried out, which helps to quantize the image for obtaining cartoon like effect. Different image processing techniques such as contrast stretching, bilateral filtering, luminance quantization, and edge detection are being implemented to achieve the desired outcome. We have proposed and implemented a device using Raspberry pi that could create cartoonized images on the fly from a camera feed. OpenCV programming is used for implementing image processing techniques.

 

Published In : IJCAT Journal Volume 1, Issue 6

Date of Publication : 31 July 2014

Pages : 284 - 287

Figures :11

Tables : --

Publication Link : Real-Time Cartoonization Using Raspberry Pi

 

 

 

Parimal Sikchi : School of Electronics Engineering (SENSE) VIT University, Vellore-632014 India

Neha Beknalkar : School of Electronics Engineering (SENSE) VIT University, Vellore-632014 India

Swapnil Rane : School of Electronics Engineering (SENSE) VIT University, Vellore-632014 India

 

 

 

 

 

 

 

Cartoonization

Bilateral Filter

Luminance quantization

Camera feed

Raspberry pi

OpenCV

We have implemented several filtering and image enhancement techniques to obtain desired effects of cartoonization. We have used the theory of bilateral filtering, to filter out the small discontinuities and strengthening the salient edges, along with unsharp masking for feature detailing. In addition to this, image frames are processed by image enhancement techniques like contrast enhancement, edge enhancement. Luminance quantization helped for getting color reduction and quantization effects. It is seen that accuracy and performance of this algorithm also depends on the resolution of the image. Some important details might be lost while cartoonifying due to poor resolution of the input image, but it is also seen that bigger the image more time it takes to form cartoonized image. Hence, there was an inevitable trade-off between these two factors. To get better performance and speed up the system, we have developed hardware system which improves efficiency.

 

 

 

 

 

 

 

 

 

[1] Chaoqun Hong, Zhi Yang, Jiajun Bu, Yi Liu, Chun Chen, Cartoon Like Stylization of video real time applications, Zhejiang University, IEEE International Conference on Multimedia and expo 978-1-4244- 2571-6/08/©2008 J.ClerkMaxwell,ATreatiseonElectricityandMagnetism ,3rded.,vol.2.Oxford:Clarendon,1892,pp.68–73.

[2] ZoyaShahcheraghi #1, John See, On the Effects of Pre- and Post-Processing in Video Cartoonization with Bilateral Filters, IEEE international conference on Signal and Image processing applications(ICSIPA),978-1-4799-0267-5.

[3] Raman Maini, Dr. Himanshu Aggarwal “Study and Comparison of Various Image Edge Detection Techniques”, International Journal of Image Processing (IJIP), Volume (3) : Issue (1). J. Canny. A Computational Approach to Edge Detection. IEEE Transactions on Pattern Analysis & Machine Intelligence,8(6):679–698,November 1986.M.Young,TheTechnicalWriter'sHandbook.Mill Valley,CA:UniversityScience,1989.

[4] H. Winnemoller, S. C. Olsen, and B. Gooch, "Real- Time Video Abstraction," ACM Transactions on Graphics, pp.1221- 1226,2006J.ClerkMaxwell,ATreatiseonElectricityand Magnetism,3rded.,vol.2.Oxford:Clarendon,1892,pp.68 –73. C. Tomasi and R. Manduchi. Bilateral Filtering for Gray and Color Images. In International Conference on Computer Vision (ICCV), pages 839–846, 1998.

[5] N. Chaaklha, W.C. Tan, Teresa H. Y. Meng: Color Quantization of Images Based on Human Vision Perception. IEEE International Conference, 1994.R.Nicole,“Titleofpaperwithonlyfirstwordcapitali zed,” J.NameStand. Abbrev.,inpress.

[6] Fast color quantization using weighted sort-means clustering. JOSA A, Vol. 26, No.11, pp.2434-244, 2009.