Implementation of 5-Block Convolutional Neural Network (CNN) for Saliency Improvement on Flying Object Detection in Videos
Shubha Pandey, Ritu Sharma, Ghanapriya Singh · 2020
In this paper an algorithm for detecting a flying object is proposed from the recorded videos using stationary as well as moving camera. Firstly the methods which required less computational power and have less complexity are used. The signature saliency which performs very well and there are very few misses of object detection is used. At the last stage only if required the CNN block is called when the false positive result from previous stages is more.