Military vehicle detection and classification

Project Title

Military vehicle detection and classification

Project Area of Specialization Artificial IntelligenceProject Summary

Project Summary 

Vehicle recognition is a crucial part of intelligent military vehicle categorization and identification. When contrasted with images of various military vehicles, which are less dissimilar in colour, it becomes more difficult to distinguish military vehicles. In this study, we present a figure identification approach based on the YOLO v4 deep learning methodology for recognizing and localizing military vehicles in tough surroundings quickly and accurately. This project compares the detection effects of YOLO v4 with those of Faster R-CNN and YOLO v3, which were previously widely used in the field of military vehicle recognition, using the same figure dataset. The testing findings show that the detection impact of the Military Vehicle Recognition Model based on the YOLO v4 algorithm has increased to some level in terms of average precision and other fundamental metrics. It shows that the YOLO v4 deep learning system can detect figures in a complex environment and provide technological support for smart figure vehicle management..

Project Objectives

Project Objectives

• Design and implementation of smart detection of military vehicle.
• Computer aided detection (CAD) systems are intended to improve accuracy by using YOLOV4 technique
• Detection of 5 objects in real time. 
   

Project Implementation Method

Project Implementation Method

First of all, create the data set of images. Upload the obj files to the googldrive and then we have to link our Collab Notebook with the Google drive and Mount that drive.

Then we run the code that will Train the Detector on Multiple iterations. Then after completing the training detector is ready to detect any image or video input in which the military vehicle is present. The camera is used to take image then label it. Compare the input images with data base. At the end display percentage output.

Benefits of the Project

Benefits of the Project

Detecting military vehicles and distinguishing them out from non-military vehicles is a significant challenge in the defense sector. Detection of military vehicle could help to identify enemy’s move and hence, build early precautionary measures. Recently, many deep learning-based techniques have

been proposed for vehicle detection purpose. However, they are developed using datasets that are not useful if military specific vehicle training and detection is required. Hyper-parameters in those techniques are not tuned to entertain low-altitude aerial imagery. We aim to develop state-of-the-art deep learning framework to detect particularly military vehicle. The major bottleneck in the application of deep learning frameworks to detect military vehicles is the lack of available datasets. In this context, we prepared a dataset of low-altitude aerial images that comprises of real data (taken from military shows videos) and toy data (taken from YouTube videos). Our dataset is categorized into five main types i.e. military_tank, Helicopter, Jeeps, Military Truck and Air crafts .

Technical Details of Final Deliverable

Technical Details of Final Deliverable 

Detection

• We have tested some images and videos meanwhile and it has been satisfactory so far.
• And also, project is tested on live webcam, on images as well as on video.

You can check the mAP for all the saved weights to see which gives the best results ( 2000 here
is the saved weight number like 4000, 5000 or 6000 and so on )

Graph Between Iterations and Average loss

Final Deliverable of the Project Software SystemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Responsible Consumption and ProductionRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	6625
Google-COLAB-Pro-Account	Miscellaneous	1	3625	3625
Report Binding	Miscellaneous	6	500	3000