Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection

As the performance of a convolutional neural network is logarithmically proportional to the amount of training data, data augmentation has attracted increasing attention in recent years. Although the current data augmentation methods are efficient because they force the network to learn multiple par...

Full description

Saved in:
Bibliographic Details
Main Authors: Huan Zhang, Xiaolin Han, Weidong Sun
Format: Article
Language:English
Published: MDPI AG 2024-12-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/24/4806
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846102832316416000
author Huan Zhang
Xiaolin Han
Weidong Sun
author_facet Huan Zhang
Xiaolin Han
Weidong Sun
author_sort Huan Zhang
collection DOAJ
description As the performance of a convolutional neural network is logarithmically proportional to the amount of training data, data augmentation has attracted increasing attention in recent years. Although the current data augmentation methods are efficient because they force the network to learn multiple parts of a given training image through occlusion or re-editing, most of them can damage the internal structures of targets and ultimately affect the results of subsequent application tasks. To this end, region-focusing data augmentation via salient region activation and bitplane recombination for the target detection of optical satellite images is proposed in this paper to solve the problem of internal structure loss in data augmentation. More specifically, to boost the utilization of the positive regions and typical negative regions, a new surroundedness-based strategy for salient region activation is proposed, through which new samples with meaningful focusing regions can be generated. And to generate new samples of the focusing regions, a region-based strategy for bitplane recombination is also proposed, through which internal structures of the focusing regions can be reserved. Thus, a multiplied effect of data augmentation by the two strategies can be achieved. In addition, this is the first time that data augmentation has been examined from the perspective of meaningful focusing regions, rather than the whole sample image. Experiments on target detection with public datasets have demonstrated the effectiveness of this proposed method, especially for small targets.
format Article
id doaj-art-0cb7c36ae2e44f1b828eef5102952d46
institution Kabale University
issn 2072-4292
language English
publishDate 2024-12-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-0cb7c36ae2e44f1b828eef5102952d462024-12-27T14:51:22ZengMDPI AGRemote Sensing2072-42922024-12-011624480610.3390/rs16244806Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target DetectionHuan Zhang0Xiaolin Han1Weidong Sun2Department of Electronic Engineering, Tsinghua University, Beijing 100084, ChinaSchool of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaDepartment of Electronic Engineering, Tsinghua University, Beijing 100084, ChinaAs the performance of a convolutional neural network is logarithmically proportional to the amount of training data, data augmentation has attracted increasing attention in recent years. Although the current data augmentation methods are efficient because they force the network to learn multiple parts of a given training image through occlusion or re-editing, most of them can damage the internal structures of targets and ultimately affect the results of subsequent application tasks. To this end, region-focusing data augmentation via salient region activation and bitplane recombination for the target detection of optical satellite images is proposed in this paper to solve the problem of internal structure loss in data augmentation. More specifically, to boost the utilization of the positive regions and typical negative regions, a new surroundedness-based strategy for salient region activation is proposed, through which new samples with meaningful focusing regions can be generated. And to generate new samples of the focusing regions, a region-based strategy for bitplane recombination is also proposed, through which internal structures of the focusing regions can be reserved. Thus, a multiplied effect of data augmentation by the two strategies can be achieved. In addition, this is the first time that data augmentation has been examined from the perspective of meaningful focusing regions, rather than the whole sample image. Experiments on target detection with public datasets have demonstrated the effectiveness of this proposed method, especially for small targets.https://www.mdpi.com/2072-4292/16/24/4806region-focusingdata augmentationsalient region activationbitplane recombinationtarget detection
spellingShingle Huan Zhang
Xiaolin Han
Weidong Sun
Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
Remote Sensing
region-focusing
data augmentation
salient region activation
bitplane recombination
target detection
title Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
title_full Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
title_fullStr Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
title_full_unstemmed Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
title_short Region-Focusing Data Augmentation via Salient Region Activation and Bitplane Recombination for Target Detection
title_sort region focusing data augmentation via salient region activation and bitplane recombination for target detection
topic region-focusing
data augmentation
salient region activation
bitplane recombination
target detection
url https://www.mdpi.com/2072-4292/16/24/4806
work_keys_str_mv AT huanzhang regionfocusingdataaugmentationviasalientregionactivationandbitplanerecombinationfortargetdetection
AT xiaolinhan regionfocusingdataaugmentationviasalientregionactivationandbitplanerecombinationfortargetdetection
AT weidongsun regionfocusingdataaugmentationviasalientregionactivationandbitplanerecombinationfortargetdetection