[This article belongs to Volume - 54, Issue - 02]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-23-10-2022-358

Kiran Phalke, Dr. R. S. Hegadi

Abstract :

This paper intends to study Algorithms created for real-time object identification applications to increase detection accuracy and energy efficiency by combining Convolutional Neural Networks (CNN) with the Scale Invariant Feature Transform. Object detection has been a research community focus for many years and has achieved substantial breakthroughs in its trip thus far. There is a vast array of applications that might benefit from greater advancements in the field of object detection. The authors of the current study implemented real-time object detection and worked to increase the detection mechanism's accuracy. We utilized the ssd v2 inception coco model in this study because Single Shot Detection models produce much superior outcomes. A dataset of more than 100 raw photos is utilized for training, and labellimg is used to produce xml files. The produced tensor flow records are sent into training pipelines that employ the suggested model. OpenCV collects real-time pictures, whereas CNN executes image convolution processes. The real-time object detection achieves an accuracy of 92.7%, which is an improvement above some of the previous models provided. The model identifies hundreds of items at the same time. The suggested methodology greatly improves on existing techniques in practice in terms of object detection accuracy. There is a large dataset available to assess the correctness of the suggested model. The model might be beneficial for a variety of item detection applications, such as parking lots, people identification, and inventory management. CNN (Convolution Neural Network) is a subset of deep learning algorithms that can take as input a sample picture and execute convolution operations to extract characteristics from the image and distinguish one object from the others. Rajaraman et al., 2019; Alganci et al., 2020). Consider tracing our misplaced phone in an untidy and cluttered residence to examine the application domain of machine learning systems. It appears to be a cumbersome and frustrating task for anyone. It needs only a few milliseconds to track the Location of mobile. Well, this is precisely the power we can harness from these amazing object detection algorithms, which are at the bottom of heart the deep learning algorithms. The current research work focuses on proposing an object detection model that can take input from the web camera, find location of the object through webcam, and classify object on screen for its appropriate category.