Improvement of the model of object recognition in aero photographs using deep convolutional neural networks

Vadym Slyusar,Serhii Vazhynskyi,Oleg Bogatov,Kirill Khrustalev,Serhii Shevchenko,Oleksandr Chernikov,Pavlo Borodych,Pavlo Kovalov,Anton Chernukha,Mykhailo Protsenko

doi:10.15587/1729-4061.2021.243094

Abstract

Detection and recognition of objects in images is the main problem to be solved by computer vision systems. As part of solving this problem, the model of object recognition in aerial photographs taken from unmanned aerial vehicles has been improved. A study of object recognition in aerial photographs using deep convolutional neural networks has been carried out. Analysis of possible implementations showed that the AlexNet 2012 model (Canada) trained on the ImageNet image set (China) is most suitable for this problem solution. This model was used as a basic one. The object recognition error for this model with the use of the ImageNet test set of images amounted to 15 %. To solve the problem of improving the effectiveness of object recognition in aerial photographs for 10 classes of images, the final fully connected layer was modified by rejection from 1,000 to 10 neurons and additional two-stage training of the resulting model. Additional training was carried out with a set of images prepared from aerial photographs at stage 1 and with a set of VisDrone 2021 (China) images at stage 2. Optimal training parameters were selected: speed (step) (0.0001), number of epochs (100). As a result, a new model under the proposed name of AlexVisDrone was obtained. The effectiveness of the proposed model was checked with a test set of 100 images for each class (the total number of classes was 10). Accuracy and sensitivity were chosen as the main indicators of the model effectiveness. As a result, an increase in recognition accuracy from 7 % (for images from aerial photographs) to 9 % (for the VisDrone 2021 set) was obtained which has indicated that the choice of neural network architecture and training parameters was correct. The use of the proposed model makes it possible to automate the process of object recognition in aerial photographs. In the future, it is advisable to use this model at ground stations of unmanned aerial vehicle complex control when processing aerial photographs taken from unmanned aerial vehicles, in robotic systems, in video surveillance complexes and when designing unmanned vehicle systems

Highlights

The new trend in the development of machine vision and artificial intelligence consists in working out a technology of automated monitoring of critical infrastructure objects using unmanned aerial vehicles (UAVs) [1]
To recognize objects in aerial photographs taken with cameras installed on UAVs, it was proposed to use deep convolutional neural networks (DCNN)
Application of the proposed model makes it possible to: ‒ automate the process of recognition of objects in aerial photographs; ‒ improve the accuracy of object recognition in images from 7 % to 9 % which confirms that the choice of the neural network architecture and training parameters was correct

Summary

Introduction

The new trend in the development of machine vision and artificial intelligence consists in working out a technology of automated monitoring of critical infrastructure objects using unmanned aerial vehicles (UAVs) [1]. Such objects include enterprises of strategic importance for the economy and security of the state [2], facilities of power. Disruption of the normal functioning of these facilities can threaten vital national interests [6,7,8,9] In this regard, it is necessary to develop artificial intelligence systems and improve methods of their implementation in unmanned aerial vehicle complexes (UAVCs) [10]. The improvement of the model of object recognition in aerial photographs using the DCNNs is an urgent problem

Literature review and problem statement

The aim and objectives of the study

The study materials and methods

The results obtained in the study of object image recognition using DCNN

Findings

Conclusions