SURVIVAL STUDY ON OBJECT DETECTION BASED CLASSIFICATION WITH AERIAL IMAGES

Abstract

Aerial imagery is the process of capturing the photos from flying entity. The aerial imagery has several objects each belonging to different classes. Object detection is more demanding one as it includes bounding box around every object in image and allocates into the class label. Object recognition denotes the group of associated tasks for detecting the objects in photographs. Image classification is defined as the process of assigning the class label to every image. Feature plays an essential part in aerial image classification for performing the object detection. Feature extraction includes the extraction of relevant shape information for performing accurate classification by formal process. Classification in aerial images is one of the key issue over last decade which categorize the image based on object exists. But, multiclass classification of objects in aerial images is still difficult one as classification accuracy was not improved. In our research work, many object detection based image classifications are studied and identified the existing problems.