Zeshan Hussain

STEP 1: Gathering Image Database; I Begin by gathering a collection of images to train the system. You will need both positive and negative samples. Positive images should contain the object you want to detect, while negative images should not include the object. You can start with how many positive and negative images you want to provide for training, the more, the accurate the model will be. However, you have the option to expand your dataset by capturing additional sequences using HAKA1 or sourcing public images from online databases. Remember, having a diverse set of positive and negative images improves the accuracy of the classifier.

STEP 2: Organizing Negative Images Next, I organized my negative images by placing them in the folder named ...\training\negative. Execute the batch file named create_list.bat, which generates a text file listing all the background images. By running the batch file, I will obtain a text file containing the names of your negative images, each listed on a separate line. This file will be used during the training process to enhance the classifier's performance. Batch File Command:

create_list.bat 2 dir /b *.jpg >bg.txt Running this batch file, you will get a text file each line looks as below: image1200.jpg image1201.jpg image1202.jpg ... Later, we need this negative data file for training the classifier.

STEP 3: Cropping & Marking Positive Images In this step, I'll prepare the positive images by marking the location of the objects within them. You can accomplish this task using one of two utilities: Objectmarker or Image Clipper. Objectmarker is the simpler and faster option, while Image Clipper provides more versatility albeit with additional time investment. For now, let's proceed with Objectmarker for its straightforward approach. Begin by placing your positive images in the folder ...\training\positive\rawdata. Within the same directory, you'll find a file named objectmaker.exe, which is essential for marking the objects in the positive images. Please ensure that the files cv.dll and highgui.dll are present in the same directory as objectmaker.exe for it to function correctly. Before launching objectmaker.exe, find a comfortable and uninterrupted environment as you'll need to meticulously mark and crop numerous images. Upon launching objectmaker.exe, you'll encounter two windows: one displaying the loaded image and the other revealing the image's filename. Follow the prompts to mark the objects accurately within your positive images.

STEP 4: Creating a vector of positive images In folder ..\training\ there is a batch file named samples_creation.bat The content of the batch file is: createsamples.exe -info positive/info.txt -vec vector/facevector.vec -num 200 -w 24 -h 24 Main Parameters: -info positive/info.txt Path for positive info file -vec vector/facevector.vec Path for the output vector file -num 200 Number of positive files to be packed in a vector file -w 24 Width of objects -h 24 Height of objects The batch file loads info.txt and packs the object images into a vector file with the name of e.g. facevector.vec After running the batch file, you will have the file facevector.vec in the folder..\training\vector Note: To run creatsample.exe you also need the files cv097.dll, cxcore097.dll, highgui097.dll, and libguide40.dll in the folder ..\training.

				
#include opencv2/opencv.hpp
#include opencv2/highgui/highgui.hpp
#include opencv2/imgproc/imgproc.hpp
#include iostream
#include opencv2/objdetect/objdetect.hpp

using namespace cv;
using namespace std;



void VideoStream() {
CascadeClassifier firearm_cascade;
    firearm_cascade.load("/Users/zximy/Desktop/Stage1Firearm.xml"); /* Haar Cascade file */

    String url = "http://192.168.1.118/640x480.jpg";

    namedWindow("SCANNING AREA", WINDOW_AUTOSIZE);

    while (true) {
        VideoCapture videoResponse(url);

        Mat videostream;
        videoResponse >> videostream;

        Mat gray;
        cvtColor(videostream, gray, COLOR_BGR2GRAY);

        vector gundetected;
        firearm_cascade.detectMultiScale(gray, gundetected, 12, 26);

        for (const Rect& rect : gundetected) {
            rectangle(videostream, rect, Scalar(0, 0, 255), 3);
        }

        imshow("Scanning", videostream);

        char key = waitKey(5);

        if (key == 'q') {
            break;
        }
    }
}
int main() {
    VideoStream();

    return 0;
}