You are on page 1of 1

OBJECT DETECTION AND RECOGNITION

ABSTRACT
Real time object detection is a vast, vibrant and complex area of computer vision.
Deep learning methods focus on useful features, which are extracted automatically by the
machine. These methods analyze a huge amount of extracted features with a logical structure
similar to the human brain, and this characteristic enables these methods to get more accurate
results compared with conventional machine learning methods. The advancement of deep
learning algorithm helps to develop object detection application easily than before. In this
research work, a deep learning-based system is developed for real-time object detection from
webcam captured videos. You Only Look Once (YOLO), which is widely used for object
detection, has been applied in this research work. For separated bounding box, it frames
object detection in images as a regression problem. It provides an image with caption and the
object is highlighted with the probability of correct detection. To contract and train the object
detection model here TensorFlow API, a machine learning open-source framework is used.

The main objectives of our research are:


(i) To detect Objects with high accuracy and high confidence,
(ii) To detect objects despite being only partially visible (occlusion),
(iii) To detect objects when present in large groups irrespective of the size of
the object.
(iv) To detect objects with low illumination.s

You might also like