For example, you can read, show and write images using the imread, imshow and imwrite commands respectively.

#include <stdio.h>
#include <opencv2\opencv.hpp>
#include <string>
#include <fstream>

using namespace std;

int main(int argc, char* argv[])
{
	// ============= Read, Show, Write Image ===================
	// CAUTION!!! This statement requires a jpg file
        //in the current directory
	cv::Mat img = cv::imread("test.jpg");

	if(img.data!=NULL){
	   cv::Mat scaledimg(img.rows/img.cols*640,640,img.type());
		cv::resize(img,scaledimg,scaledimg.size());

		cv::namedWindow("Image");

		cv::imshow("Image",scaledimg);
// See what happens if you don't use this
		cv::waitKey();
		cv::imwrite("test_scaled.jpg",scaledimg);
	}
	return 0;
}

When the question of video comes, OpenCV has a one stop solution named VideoCapture. It can capture video frames from both a video or a webcam.

#include <stdio.h>
#include <opencv2\opencv.hpp>
#include <string>
#include <fstream>

using namespace std;

int main(int argc, char* argv[])
{
	// ============== Capture Camera Frame ====================
	cv::Mat frame;
	// Capturing data from camera
        // Caution: It requires webcam to be attached
	cv::VideoCapture cap(0);
        // you can also use
        // cv::VideoCapture cap("video filename");
        // to capture the frame from a video instead of webcam
	if(!cap.isOpened())
	  printf("No Camera Detected");
	else{
		cv::namedWindow("Webcam Video");
		for(;;){
			cap >> frame; // get a new frame from camera
			cv::imshow("Webcam Video",frame);
			if(cv::waitKey(30) >= 0) break;
		}
	}
	return 0;
}

Note: This post assumes you have OpenCV 2.3 installed. If not, please check the following post
http://www.itanveer.com/2011/installing-opencv-231/

// OpenCV_Test.cpp : Defines the entry point for the console application.
//

#include <stdio.h>
#include <opencv2\opencv.hpp>
#include <string>
#include <fstream>

using namespace std;

// I wrote this method to display matrix contents. It is only for
// double (64 bit) data type. I shall write a more general version
// in near future
void dispMat(CvMat *N, string varName){
    double *data = N->data.db;
    int rowNum = N->rows;
    int colNum = N->cols;

    printf("\n========= Matrix Display System ==============\n");
    printf("(%d,%d) %s \n",rowNum,colNum,varName.c_str());
    for(int i=0;i<rowNum;i++){
        if(colNum>10)
            printf("Row No=%d\n",i+1);
        for(int j=0;j<colNum;j++)
            printf("%0.2f\t",data[i*colNum+j]);
        printf("\n");
    }
    printf ("=================================================");
    printf("\n");
}

int main(int argc, char* argv[])
{
    int a=0;

    double c[] = {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16};
    double d[] = {2,1,6,-4,0.5,4,9,1,9,-1,-0.3,0.7,3,2,1,1.5};

    // Declaring Matrix as a Pointer and initialization
// This is the old style of OpenCV Matrices
    CvMat* M = cvCreateMat(4,4,CV_64FC1);
    cvInitMatHeader(M,4,4,CV_64FC1,d);
    dispMat(M,"M = ");

    // Declaring Matrix as variable and initialization
    CvMat N = cvMat(4,4,CV_64FC1,c);
    dispMat(&N,"N = ");

    // Declaring Matrix as pointer without initialization
    CvMat* M2;

    //Cloning a Matrix
    M2 = cvCloneMat(M);
    dispMat(M2,"M2 = ");

    //indirect access to matrix element
    //get
    printf("\n(2,3)th element in M2 = %0.2f\n",cvmGet(M2,2,3));
    //set
    cvmSet(M2,2,3,90);
    printf("(2,3)th element is now set to = %0.2f\n",
	cvmGet(M2,2,3));

    cout<<"\nPress Enter to continue\n";cin.get();
    // ================================================

    //Set zero
    cvSetZero(M2);
    dispMat(M2,"\nM2 (cvSetZero)\n");

    //Set Identity
    cvSetIdentity(M2);
    dispMat(M2,"\nM2 (cvSetIdentity)\n");

	cout<<"\nPress Enter to continue\n";cin.get();

    // ============ OpenCV 2 Test =====================
    cv::Mat A = cv::Mat(M,true);
    cv::Mat J = cv::Mat(4,1,CV_64FC1,cv::Scalar(1));

    cv::Mat B = (J.t())*(A);

    CvMat AA = A;
    dispMat(&AA,"OpenCV 2 Variable: A");

    AA = J;
    dispMat(&AA,"OpenCV 2 Variable: J");

	// Try this: B  = (J.t()*A).t();
    B  = (J.t()*A);
	B = B.t();
	AA = B;		// Also try this: AA = B.t();
    dispMat(&AA,"OpenCV 2 Variable:(J'*A)'");

    cout<<"\nPress Enter to continue\n";cin.get();
    // =========== Predefined openCV 2 Matrices ========
    A = cv::Mat::zeros(5,2,CV_64FC1);
    AA = A;
    dispMat(&AA,"Predefined matrices: zeros(5,2,cv_64fc1)");

    A = cv::Mat::ones(5,2,CV_64FC1);
    AA = A;
    dispMat(&AA,"Predefined matrices: ones(5,2,cv_64fc1)");

    A = cv::Mat::eye(5,5,CV_64FC1);
    AA = A;
    dispMat(&AA,"Predefined matrices: eye(5,5,cv_64fc1)");

    cout<<"\nPress Enter to continue\n";cin.get();
    // =========== Predefined openCV 2 Matrices ========
    printf("\n\n Element (1,2) = %f",A.at<double>(2,1));

    // === Writing opencv Matrix (M2) to and from Filestream =====

    ofstream ofs("d:\\test.dat",ios::out|ios::binary);

    dispMat(M2,"Matrix before writing to file");

    double p =5;int s=3;
    double* m2Dat = M2->data.db;

    ofs<<p<<endl<<s;
    ofs.write((char *) m2Dat,M2->cols*M2->rows*sizeof(double));

    ofs.close();

    p=0;s=0;
    m2Dat[0]=-1;
    m2Dat[1]=-1;
    m2Dat[2]=-1;
    m2Dat[3]=-1;
    m2Dat[4]=-1;
    m2Dat[5]=-1;
    m2Dat[6]=-1;
    m2Dat[7]=-1;
    m2Dat[8]=-1;
    m2Dat[9]=-1;
    m2Dat[10]=-1;
    m2Dat[11]=-1;
    m2Dat[12]=-1;
    m2Dat[13]=-1;
    m2Dat[14]=-1;
    m2Dat[15]=-1;

    dispMat(M2,"Matrix after changing data");

    ifstream ifs("d:\\test.dat",ios::in|ios::binary);

    ifs>>p>>s;
    printf("%0.3f\n",p);
    printf("%d\n",s);
    ifs.read((char *)m2Dat,M2->cols*M2->rows*sizeof(double));

    dispMat(M2,"Matrix after Loading again");

    CvMat* M3 = cvCreateMat(4,4,CV_64FC1);
    cvInitMatHeader(M3,4,4,CV_64FC1,m2Dat);
    dispMat(M3,"Matrix reading from file");

    ifs.close();

    // Releasing Matrix Handle
    cvReleaseData(M);
    cvReleaseData(M2);
    cvReleaseData(M3);

    return 0;
}

Note: This post assumes you have OpenCV 2.3 installed. If not, please check the following post
http://www.itanveer.com/2011/installing-opencv-231/

Preparation Step: Install Visual Studio 10.

Stage One: Installing OpenCV

1. Download OpenCV 2.3.1 Superpack from the following link: http://sourceforge.net/projects/opencvlibrary/files/opencv-win/2.3.1/.
2. Run the exe file and extract the contents in C:\OpenCV231. If the contents (i.e. the folders named 3rdparty, android etc) are copied in a folder named “opencv” inside C:\OpenCV231 then copy all of them to C:\OpenCV231. (I’ll describe all the later steps assuming that the files has been extracted here. If not, please modify the instruction accordingly.)
   

   Figure 1: Selecting the Generator

   

   Figure 2: cmake window

3. Download and install cmake from http://www.cmake.org/
4. Run the Graphical User Interface of Cmake (cmake-gui).
5. In the field named “Where is the source code:” white c:\opencv231. In the field named “Where to build the binaries” specify C:\OpenCV231\msvc. Now click “configure”.
6. A dialog box will appear (Figure 1) asking you to specify the generator for the project. Select Visual Studio 10 from the dropdown list. Select “Use default native compiler”. Now click “Finish”. It may ask for your permission to create a new folder named msvc. Just allow to do it. It
7. Now the cmake window will look something like Figure 2. In this dialog, check the box named “install_c_examples”. Click the “Generate” button. This will make Visual Studio solutions in the C:\OpenCV231\msvc folder.
8. Open OpenCV.sln. Select “Debug” in the solution configuration listbox. Now press F7 to build the solution. If all the projects are successfully built (3 projects may be skipped, don’t worry about that) it will show such message. Now do the same by selecting “release” in the solution configuration listbox.
9. Copy the “lib” and “bin” folders from C:\OpenCV231\msvc to C:\OpenCV231. Also copy the contents of “include” folder from C:\OpenCV231\build to C:\OpenCV231\include
10. Now we need to add the following 2 folders in environment variables:
    C:\OpenCV231\bin\Release
    C:\OpenCV231\bin\Debug
    To do this, go to Control Panel > System > Advanced System Settings > Environment Variables and append these two folders into the “path” variable. Remember to separate the entries by semicolon. Click ok.

Compatibility Issue:

I have found the following backward compatibility issue while moving from opencv 2.1 to 2.3.1

If M be a cv::Mat then if you have a command M.inv(CV_SVD_SYM) then you have to change it to M..inv(cv::DECOMP_SVD)

The Use of Active Appearance Models

With Professor Jeffrey Cohn

Renowned researcher of psychology Professor Dr. Jeffrey Cohn, in his presentation “Machine Learning for Affective Computing: Findings and Issue” showed how Active Appearance Model has paved the way for some of the experiments which were not possible to conduct earlier. For example, the capability of Active Appearance Models to accurately model and tack a person’s head movement and facial expressions is utilized to generate almost real time (~0.5 sec delay) avatars. These avatars are used in interaction between two people in several psychological experiments. For example, it was shown that, people can interact as easily with an avatar as a normal human being. However, these interactions can be manipulated by reducing the amount of expressions (which can be done by computationally reducing the expressions expressed by the avatars) or by immobilizing the head movement or by changing the gender of the avatar (A male showing a female avatar and vice versa) etc. In those cases, people do not identify such interactions as spontaneous human-human interactions. Prof Cohn also showed the effect of smile dynamics of faces i.e. how the “spontaneity” in smile is related with the temporal evolution. In a later presentation named “FAST-FACS: A Computer-Assisted System to Increase Speed and Reliability of Manual FACS Coding” Jeff Cohn also showed how the manual landmark annotation effort can by reduced by 40% through the use of some semi-automatic annotation process.

Human centric Machine Learning

Some exciting achievements are possible to make by incorporating humans in different aspects of machine learning. This point is underscored by the invited speaker from Microsoft research, Ashish Kapoor, in his presentation “Human-Centric Machine Learning”. He mentioned that, it is possible to use human interactions more than just data labeling. He showed several of his works on how it is possible to use human interactions to rectify the decision boundaries of classifiers, incorporate experience and use in many other exciting ways.

More on Smile Dynamics

More research on the types of smile has been shown in the paper “Are You Friendly or just Polite? – Analysis of smiles in spontaneous face-to-face interactions” presented by Mohammed Hoque, Louis-Philippe Morency and Rosaline Picard. It has been shown that most of the spontaneity information lies on what time it takes for a smile to rise to the peak from the onset and decay to the offset from the peak. Mohammed Hoque mentioned that MIT media lab will release the dataset that has been used in this work. I believe it will give us a rare opportunity to access some spontaneous dataset just for free.

Interactions and Socialization

With Professor Rosaline Picard, the founder of Affective Computing

This conference gave me a wonderful opportunity to see and discuss with the leading researchers in affective computing. I talked personally with Jeffry Cohn, Rosaline Picard, Peter Robinson, Louis-Philippe Morency and asked few questions regarding current trends of research in affective computing. In a conversation with Prof Cohn on the channels of emotion, he mentioned that he is a bit skeptical on how far it is possible to infer emotion appropriately from physiological channels. He mentioned two reasons for that. Firstly, physiological responses are relatively slower than the response showed in facial expressions. Secondly, sensors required to measure physiological signals can alter the emotions to some extent.

Professor Rosaline Picard highly appreciated the project concept of blind emotion. She mentioned that it is possible to help blind people in great extent to perceive others emotion. MIT Media lab did some preliminary investigations on these issues. She specifically pointed that it will not be a good idea to convey the facial expressions (like smile, frown etc.) to the people who are congenitally blind. Rather, it is better to convey some inferred information which makes more sense to them. For example, whether other people are interested on what a blind person is speaking etc. Prof Picard also pointed on a possible difficulty that their lab faced. She mentioned that it becomes very difficult to track and analyze a person’s face with the inherent head movement of the wearer. She also welcomed any help and collaboration that may be asked from our side.

Dr. Louis-Philippe Morency showed an extra ordinary enthusiasm to our effort on analyzing

With Professor Peter Robinson

facial expressions and works with active appearance model. He mentioned that he is going to release a more robust version of Constrained Local Model (CLM) based face tracker. This tracker will be released in public domain. He also expressed his ideology on how research related data and tool should be disseminated for public use. He also gave word to entertain our request for this face tracker as much as possible.

It is apparent from a long growing mass of evidences that the part of a brain responsible for processing information from a particular sensory channel gets recruited for other activities after loss of that particular channel. This phenomenon is referred to as “neuroplastic” behavior of brain. When it contributes to the improvement of other sensory channels, it is called “crossmodal neuroplasticity”. This paper refers to a considerable amount of reported evidences on crossmodal neuroplasticity following sensory deprivation and came to some conclusions based on those observations.

It particularly discussed about the following aspects of neuroplasticity. Firstly, it refers to a significant amount of reported experiments and their results on the effect of loss of vision and auditory capabilities. Analyzing these experiments, a conclusion was made that the loss of a sensory channel translates into an improved or at least an equal performance in behavioral and sensory tasks. Experimental proofs have been referred in this paper where it is reported that blind individuals showed equal or superior performance in tactile-discrimination thresholds, auditory-pitch discrimination, spatial sound localization, speech discrimination, verbal recall etc. In addition, it has been discussed that electroencephalographic and neuro-imaging technologies provide strong evidence on an altered association of brain areas for proficient utilization of other sensory channels. Similar works have also been showed for deafness and loss of multiple sensory channels.

Other than collecting evidences of crossmodal neuroplasticity, this paper also discussed on reported experiments which can explain the potential causes and mechanisms under this phenomenon. A potentially viable cause of recruitment of cortical areas followed by a sensory loss is thought to be the direct connection of the neurons to the intact sensory area. This paper also discussed on the endeavors made for determining the critical age for crossmodal plasticity. The recent experiments showed that neuroplasticity can occur in a much later age than that were presumed earlier.

Lastly, the paper discussed that in spite of the compensatory behavioral gain, the neuroplasticity cannot be viewed as universally adaptive. It may also have unintended consequences. Following the rehabilitative training, the neuroplastic changes can undermine the ability of the cortex to return back to its primary function.

Amir Amedi 1,2, William Stern 1, Joan A. Camprodon 1, Felix Bermpohl 1,3, Lotfi Merabet 1, Stephen Rotman 1, Christopher Hemond 1, Peter Meijer 4 and Alvaro Pascual-Leone

This paper reports some functional characteristics of a part of brain known as the lateral-occipital tactile-visual (LOtv) area. Traditionally, this part is known to be activated when a person observes the shape of an object through vision or touch. In this paper, it is claimed that this part of brain is responsible for analyzing shape of an object regardless of the modality of the incoming signal. This assertion has been made based on an experiment where the authors analyzed the functional characteristics of the LOtv area in response to sensory substitution soundscape.

VOICE is a computer program that converts the raw pixel intensities of a picture into a soundscape. There are some blind and sighted expert users who can perceive the shapes of objects by listening to the soundscapes. Neural images of these people’s brain reveal that when they try to perceive the shape of an object from the soundscape, their LOtv area activates. Moreover, this happened only to the people who were trained to interpret the soundscape and can extract shape information from it. People, who associate a particular soundscape to a particular shape of an object, do not display any activity in this area.

The authors have proposed a hypothesis based on this observation. They claimed that, the LOtv processes the shape information regardless the modality of sensory signal.

After the Iraq war, veterans who lost there limbs, sometimes felt unnatural sensations in their lost limbs. How would you explain such phenomenon? For a long time, it was considered as a hoax. Neuroscientists could not accept this as a true sensation because for every sensation there should have some neurons carrying small electric signals to the brain. In the case of phantom limb, this is not possible. Then what?

Dr. Ramachandran made a pioneering work in explaining phantom limb syndrome while providing treatment to one of his patients named Derick Sting. Derick used to feel vivid presence of his left arm in a strange way. When he used to go for a shave, he felt sensation in the left arm which actually does not exist. It was found in some simple experiment that when Derick’s left cheek is touched, he feels that sensation both in his cheek and in the left arm. Dr. Ramachandran got very interested. He knew that different parts of our body are mapped to some part of our brain. For example, the entire left part of the body is mapped to a strip on the right part of the brain known as somatosensory cortex. So he hypothesized that, when a limb gets amputated it looses its connection from brain and that part of the brain does not receive any signal anymore. Now, without any signal, this part becomes thirsty for some sensory signal and gets more and more sensitive to any kind of electric impulses. When some electronic impulses come to the nearby region, it actually catches that signal and gives a false signal to brain regarding a sensation on the person’s lost limb. Therefore, the postulation was the brain undergoes a massive rewiring after the amputation. Dr. Ramachandran got some scornful comments regarding his hypothesis. One strong argument was based on the dogma that the brains wiring is permanent and does not change once it is developed in the fetus. However, Ramachandran’s hypothesis got support when neural images provided support on his side.

Amputation can alter brain’s work not only in receiving signal but also transmitting. A patient used to feel extreme pain like clinching his right hand too much. But there is no right hand in his body. Ramachandran explained this phenomenon like this: When we clinch our hand, it actually sends some feedback to the brain which damps brains signal to clinch. But if the limb is amputated, it does not give any feedback to the brain and as a result brain sends more and more signal which eventually leads to extreme pain.

This problem was solved by the use of a technique named as “mirror-box“. A mirror box is a kind of apparatus where you see the image of your one hand in place of your another hand. Using this box, if a person moves his left hand, he gets an illusion that his right hand is moving. The patient was advised to put both his hands into the mirror-box and move both hands. Through this, he got a visual clue that his hand his moving. This visual information also acted as a feedback to the brain and the pain started to subside. This proved Ramachandrans theory and a very intriguing assertion that even a pain can be a construct of mind.

Interesting Follow Ups

1. Ramachandran, V. S., Brang, David and McGeoch, Paul D.(2009) “Size reduction using Mirror Visual Feedback (MVF) reduces phantom pain“, Neurocase, 99999:1 DOI 10.1080/13554790903081767
2. Ramachandran, VS & Hirstein, W (1998). The perception of phantom limbs: The D. O. Hebb lecture. Brain, 121, 1603-1630.

Just imagine the money you’ve spent in your village. What will happen if few people just use it to live their lazy life? It will not only be considered as a very small amount of money but also it will be contributed to STOP the natural activity (and thoughts) that were supposed to come from those people if they would not get the money. However, if this would be used to start a little capital needed to start a small business, chances are there that one day the business would grow and grow and ignite the activity of more and more people and pull more people out of poverty (and give them opportunity to think).

I believe people really don’t need the rice prepared for them. They have every potential to earn the rice — what they need is just to get acquainted that they have the potential. That is the job what Grameen Bank does. That is why I believe in Micro-credit. I am very proud that the person who gave this concept is from the same country that I am from.

So I decided to write my Masters thesis is latex and in the process I wanted to make a latex template according to the University of Memphis requirements. After many grueling hours of handling the core of latex, finally I have completed my thesis which is now accepted by the graduate school. So, I assume a workable basic of latex class and template has been successfully created. I am uploading these files and hope that future students of the University of Memphis will find it useful.

If you improve the template and wish to share among others, I’ll be very happy to get a link of your updated files. You can use the commenting feature below to let us (me and other users) know about your modification. Please upload it somewhere in your disk-space (Dropbox, maybe) because I don’t have much space. Also, please mention what improvements have you made so that everybody gets clear about what they are downloading.

Click Here to Download