Anda di halaman 1dari 22

GESTURE

RECOGNITION
Presented By:
Ankit Mishra (10/IEE/009)

Supervised by:
Dr. Shabana Urooj
Co-Supervisor:
Prof Omar Farooq

INTRODUCTION
If development of sign language is the first step, then development of hand

recognition system using computer vision is the second step.


Gesture has been found as the best method to interact with the computer

without using other peripheral devices, such as keyboard, mouse.

GESTURE MODELLING
There are two methods used for modelling of gesture ie Temporal and Spatial.
Temporal modelling includes techniques like Hidden Markov Model, Kalman Filter,

Finite State Machine and Neural Networks etc, whereas Spatial modelling
technique takes into consideration feature like Shape, Templates, Colour of skin
or marker etc.
Modelling of shape parameter can further be classified as Geometric and NonGeometric techniques
The next step includes the processing of modelled data, here process like Noise

reduction and Normalization of data takes place


Specific feature is then extracted from the enhanced modelled data, which is

then required to recogonize the Gesture.

Fig : My Progress of work.

SPECIFICATION OF IMAGES
Name

Size
288x352x3

Bytes

Class

304128

Attributes

uint8

A = Filename: [1x48 char]

FileModDate: '10-Feb-2014 20:35:36

FileSize: 8404

Width: 352

Height: 288Bit

ColorType: 'truecolor

Name

Size

288x352

CodingMethod: 'Huffman

Bytes
101376

Class Attributes
uint8

Format: 'jpg'
Depth: 24
CodingProcess: 'Sequential'

Fig: Eg. from database

GREYSCALE IMAGE
Greyscale digital image is an image in which the value of each pixel is a

single sample, that is, it carries only intensity information. RGB images are
composed of three independent channels for red, green and blue primary
color components.

EDGE DETECTION
Areas with strong intensity contrasts

Several edge detector operators are there for generating gradient images

like sobel, prewitt, laplacian and laplacian of Gaussian (LoG).

The sobel edge detector has 3X3 neighborhood.

Prewitt operator edge detection masks are the one of the oldest and best

understood methods of detecting edges in images.

This detector finds edges by looking for zero crossings after filtering f(x, y)

with a Laplacian of Gaussian filter.

The following neighbourhood was used in my work to obtain the proper

edges of the images.

>> J = roifilt2(H, I, BW)


>> kk = im2bw(I, level) Here the threshold value best

suited was 0.3, obtained by trial and error method.

FEATURE EXTRACTION
In this work a new Algorithm has been proposed for feature

extraction.

It is based on fact that as the number of fingers used to make

a gesture increase the perimeter of hand will also increase,


therefore the gesture made can be recognized on the data
obtained from the edges of the BW image obtained above.

This value is saved for all the images in database.


Image is then classified to different class of gestures, using

supervised Learning.

RESULT
The Images Processed are classified into different classes of gestures and

then further the total perimeter of each gesture is saved as database in


their respective classes. In preceding step Query Image is requested from
the user and its features are matched against the image features which are
saved in the database.

The following is the case of wrong recognition / or case of mismatch.

CONCLUSION

From the above table it is observed that the accuracy of the system is 89%,
i.e. there are 11 cases where there is a mismatch between the image input
and the class to which it is recognized.

FUTURE WORK
In the future work, I wish to improve my work by using colour markers for

the images, and also check if the miss match in the images can be
improved by altering the orientation of Image.

REFERENCES
S. Mitra, and T. Acharya. (2007). Gesture Recognition: A Survey IEEE

Transactions on systems, Man and Cybernetics, Part C: Applications and


reviews, vol. 37 (3), pp. 311- 324, doi:10.1109/TSMCC.2007.893280.

A. Sarkar, G.Sanyal and S.Majumder (2013). Hand Gesture Recognition

Systems: Survey International Journal of Computer Applications. Volume


71-No 15, May 2013.

Rafiqul Zaman and Noor Ibraheem (2012). Hand Gesture Recognition:

Review International Journal of Artificial Intelligence & Applications (IJAIA),


Vol 3, No 4, July 2012.

Mahmoud E., Ayoub A., Jorg A., and Bernd M., (2008). Hidden Markov

Model-Based Isolated and Meaningful Hand Gesture Recognition, World


Academy of Science, Engineering and Technology 41.

THANK YOU!!

Anda mungkin juga menyukai