Dnn Based Cyber Eye for Real Time Object Detection
Abstract
The proliferation of artificial intelligence paves a way to realize real-time object detection for many computer vision applications. Deep learning technology makes it possible to achieve efficient and accurate object detection, but with large datasets and highly efficient algorithms. This work presents a new approach that converts the format of the given dataset into an HDF file format which contributes the detection algorithm to be simple, accurate, and efficient. The convolution neural network (CNN) model is trained to identify the objects present in an image and classify them. The YOLO algorithm in combination with the TensorFlow framework proves to be better in processing speed and accuracy. This model is capable of detecting multiple objects in a single frame. The simulation results show the processing time for the proposed approach is significantly reduced as compared to the existing models.