6670

Face recognition has been the core area of research in pattern recognition and machine learning during the past
two decades. Researchers around the world have proposed many techniques based statistical learning. Many
biometric systems employ face recognition to identify persons.
In this term project, we want to recognize face in an image using the statistical learning methods studied in the
course. We want to apply principal component analysis for dimensionality reduction and k-nearest neighbor
classifier for classification purpose.
Database Collection
We have collected a database of containing a total of n = 380 examples of face. There are 10 persons in this
database and 38 different views (frontal, profile, inclined …) of each person have been taken. In order words,
there are 38 images of each person in the database. The size of each image is 38 x 38 pixels (or 1444 dimensions)
stored in vector format in a mat file. Each image is a gray scale 8-bit image (i.e. no color information is available).
We do not want to extract features from the images rather we intend to use the image as feature vector. This
means that the dimension of feature space is 1444 (=d=38×38). The label of each image i.e. person id (also called
the category or class) is provided with the dataset. The images are stored in a matrix data (of size d x n) and
labels are stored in label vector (of size 1 x n): a row vector containing person id from 1 to 10. The database
can be loaded using the command load face_database.mat. Document Preview:

TERM PROJECT EE767 – Pattern Recognition Introduction Face recognition has been the core area of research in pattern recognition and machine learning during the past two decades. Researchers around the world have proposed many techniques based statistical learning. Many biometric systems employ face recognition to identify persons. In this term project, we want to recognize face in an image using the statistical learning methods studied in the course. We want to apply principal component analysis for dimensionality reduction and k-nearest neighbor classifier for classification purpose. Database Collection We have collected a database of containing a total of n = 380 examples of face. There are 10 persons in this database and 38 different views (frontal, profile, inclined …) of each person have been taken. In order words, there are 38 images of each person in the database. The size of each image is 38 x 38 pixels (or 1444 dimensions) stored in vector format in a mat file. Each image is a gray scale 8-bit image (i.e. no color information is available). We do not want to extract features from the images rather we intend to use the image as feature vector. This means that the dimension of feature space is 1444 (=d=38×38). The label of each image i.e. person id (also called the category or class) is provided with the dataset. The images are stored in a matrix data (of size d x n) and labels are stored in label vector (of size 1 x n): a row vector containing person id from 1 to 10. The database can be loaded using the command load face_database.mat. Display a Face Image from the Database th thEach image is stored in vector format. To display the i (say 60 ) image from the database, we can use the following commands on MATLAB/OCTAVE console. >> im = data(:, 60); //extract the 60 column of the matrix data >> im = reshape(im, 38, 38); //reshape the image vector to image format >> figure; imshow(im); //display the image in a figure…

Attachments:

Save your time - order a paper!

Get your paper written from scratch within the tight deadline. Our service is a reliable solution to all your troubles. Place an order on any task and we will take care of it. You won’t have to worry about the quality and deadlines

Order Paper Now

Term-Project-….pdfDataBase-MyCo….zip