GPS拒止环境巡飞器双源地理位置知觉实例分割

doi:10.7527/S1000-6893.2025.32120

电子电气工程与控制

本期目录 | 过刊浏览 | 高级检索

前一篇 | 后一篇

GPS拒止环境巡飞器双源地理位置知觉实例分割

宫冕, 张印辉(), 何自芬, 陈光晨, 张瑞

昆明理工大学机电工程学院，昆明 650500

收稿日期:2025-04-15 修回日期:2025-04-24 接受日期:2025-06-06 出版日期:2025-06-30 发布日期:2025-06-16
通讯作者: 张印辉 E-mail:zhangyinhui@kust.edu.cn
基金资助:
国家自然科学基金(62061022);国家自然科学基金(62171206);装备智能运用教育部重点实验室开放基金(AAIE-2023-0203);昆明理工大学分析测试基金(2024P20231103003)

Instance segmentation of dual-source geolocation perception for GPS reject environmental aerial vehicle

Mian GONG, Yinhui ZHANG(), Zifen HE, Guangchen CHEN, Rui ZHANG

Mechanical and Electrical Engineering，Kunming University of Science and Technology，Kunming 650500，China

Received:2025-04-15 Revised:2025-04-24 Accepted:2025-06-06 Online:2025-06-30 Published:2025-06-16
Contact: Yinhui ZHANG E-mail:zhangyinhui@kust.edu.cn
Supported by:
National Natural Science Foundation of China(62061022);The Open Fund Project of the Key Laboratory of Intelligent Application of Equipment of the Ministry of Education(AAIE-2023-0203);Analysis and Testing Foundation of Kunming University of Science and Technology(2024P20231103003)

摘要/Abstract

摘要：

针对巡飞器导航系统GPS拒止环境下无法定位及其航拍图像地物目标与背景混杂且目标边缘模糊导致分割精度低的问题，提出一种双源地理位置知觉地物目标实例分割模型（DGISM）。首先，设计沿水平、竖直和通道维度的三重自适应中心矩模块同时引入并行路径跨层交互结构聚合浅层细节信息，以提高实例特征区分度从而克服背景混杂问题。其次，设计多核空间聚焦模块融合跨尺度空间特征信息和深层通道信息以增强模型对边缘特征的聚集能力，改善了因边缘模糊导致分割精度低的问题。最后，针对GPS拒止环境下无法定位的问题，设计航拍图像与卫星信息双源感知定位模块，经分割权重提取巡飞器航拍图像特征信息与具有卫星定位信息的图像特征记忆库相匹配获取对应地理位置信息。实验结果表明：提出的双源地理位置知觉地物目标实例分割模型mAP50-95和mAP50分割精度分别达到69.1%和88.8%，图像特征双源匹配定位精度达到95.6%，在GPS拒止环境下通过分割建筑类地物目标实现巡飞器地理位置知觉精确定位。

关键词: 实例分割, 地理位置知觉, 双源定位, GPS拒止, 中心矩感知

Abstract:

To address the issues of inability to locate in GPS reject environments for the navigation system of aerial vehicles and the low seg-mentation accuracy caused by blurred edges and mixed backgrounds in aerial imagery， this paper proposes a Dual-source Geo-aware Instance Segmentation Model for ground objects， referred to as DGISM. First， a triple adaptive central moment module along horizontal， vertical， and channel dimensions is designed， along with a parallel path cross-layer interaction structure to aggregate shallow-level details， thereby enhancing instance feature discriminability and overcoming background clutter issues. Second， a multi-core spatial focus module is developed to integrate cross-scale spatial features and deep channel information， improving the model’s ability to concentrate on edge features and addressing the problem of low segmentation accuracy due to edge blur. Lastly， to tackle the problem of localization in GPS reject environments， a dual-source perception and positioning module for aerial imagery and satellite information is designed. This module extracts segmentation weights to match the feature information of the aerial vehicle’s imagery with a feature memory bank of images containing satellite positioning information， thereby obtaining corresponding geographic location information. Experimental results demonstrate that the proposed dual-source geo-aware instance segmentation model achieves mAP50-95 and mAP50 segmentation accuracies of 69.1% and 88.8%， respectively， with a dual-source feature matching positioning accuracy of 95.6%. Realizing precise geographic location perception and positioning of patrol aircraft by segmenting building targets in GPS reject environments

Key words: instance segmentation, geolocation perception, dual-source localization, gps reject, central moment perception

中图分类号:

V249

宫冕, 张印辉, 何自芬, 陈光晨, 张瑞. GPS拒止环境巡飞器双源地理位置知觉实例分割[J]. 航空学报, 2026, 47(2): 332120.

Mian GONG, Yinhui ZHANG, Zifen HE, Guangchen CHEN, Rui ZHANG. Instance segmentation of dual-source geolocation perception for GPS reject environmental aerial vehicle[J]. Acta Aeronautica et Astronautica Sinica, 2026, 47(2): 332120.

图/表 15

图1

图2

图 3

图 4

表1

表2

表3

表4

表5

表6

表7

表8

图 5

图 6

表9

参考文献 32

[1]	屈若锟，王致远，刘晔璐，等．面向城市空中交通的无人机视觉定位技术研究［J］．航空学报， 2025， 46（10）： 531168．
	QU R K， WANG ZY， LIU Y L， et al. Research on UAV visual positioning technology for urban air mobility［J］. Acta Aeronautica et Astronautica Sinica， 2025， 46（10）： 531168 （in Chinese）
[2]	HE K， GKIOXARI G， DOLLÁR P， et al. Mask R-CNN［C］∥Proceedings of the IEEE International Conference on Computer Vision. Piscataway： IEEE Press， 2017： 2961-2969.
[3]	CHEN K， PANG J， WANG J， et al. Hybrid task cascade for instance segmentation［C］∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2019： 4974-4983.
[4]	CHENG B， MISRA I， SCHWING A G， et al. Masked-attention mask transformer for universal image segmentation［C］∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2022： 1290-1299.
[5]	CHEN X， GIRSHICK R， He K， et al. Tensormask： A foundation for dense object segmentation［C］∥Proceedings of the IEEE/CVF International Conference on Computer vision. Piscataway： IEEE Press， 2019： 2061-2069.
[6]	REDMON J， DIVVALA S， GIRSHICK R， et al. You only look once： Unified， real-time object detection［C］∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2016： 779-788.
[7]	WANG X， ZHANG R， KONG T， et al. Solov2： Dynamic and fast instance segmentation［J］. Advances in Neural Information Processing Systems， 2020， 33： 17721-17732.
[8]	ZHOU C. Yolact++ better real-time instance segmentation［M］. California： University of California， 2020： 1108-1121.
[9]	BOLYA D， ZHOU C， XIAO F， et al. Yolact： Real-time instance segmentation［C］∥Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE Press，2019： 9157-9166.
[10]	DONG C， TANG Y， ZHANG L. GHA-Inst： A real-time instance segmentation model utilizing YOLO detection framework［J］. Cluster Computing， 2024， 27（6）： 7401-7415.
[11]	WANG C Y， BOCHKOVSKIY A， LIAO H Y M. YOLOv7： Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors［C］∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2023： 7464-7475.
[12]	姚俞成，李旭，徐启敏，等．基于特征增强和校准的航拍车辆实例分割方法［J］．航空学报，2023， 44（24）： 328397．
	YAO Y C， LI X， XU Q M， et al. Instance segmentation for vehicle in UAV aerial images based on feature enhancement and calibration［J］. Acta Aeronautica et Astronautica Sinica， 2023， 44（24）： 328397 （in Chinese）.
[13]	CHEN Y， YUAN X， WANG J， et al. YOLO-MS： Rethinking multi-scale representation learning for real-time object detection［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2025， 47（6）： 4240-4252.
[14]	WU H， ZHANG M， HUANG P， et al. CMLFormer： CNN and multi-scale local-context transformer network for remote sensing images semantic segmentation［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing， 2024， 17： 7233-7241.
[15]	CHEN N H， LOU T S， ZHAO L Y. MMHCA： Multi-feature representations based on multi-scale hierarchical contextual aggregation for UAV-view geo-localization［J］. Chinese Journal of Aeronautics， 2025，38（6）： 103242.
[16]	曹正阳，张冰，白屹轩，等. GNSS/INS/VNS组合定位信息融合的多无人机协同导航方法［J］.兵工学报， 2023， 44（S2）： 157-166.
	CAO Z Y， ZHANG B， BAI Y X， et al. GNSS/INS/VNS integrated positioning information fusion method for multi UAV cooperative navigation［J］. Acta Armamentarii， 2023， 44（S2）： 157-166 （in Chinese）
[17]	SUN K， MOHTA K， PFROMMER B， et al. Robust stereo visual inertial odometry for fast autonomous flight［J］. IEEE Robotics and Automation Letters， 2018， 3（2）： 965-972.
[18]	NASSAR A， AMER K， ELHAKIM R， et al. A deep CNN-based framework for enhanced aerial imagery registration with applications to UAV geolocalization［C］∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. Piscataway： IEEE Press， 2018： 1513-1523.
[19]	CHOI J， MYUNG H. BRM localization： UAV localization in GNSS-denied environments based on matching of numerical map and UAV images［C］∥2020 IEEE/RSJ International Conference on Intelligent Robots and Systems （IROS）. Piscataway： IEEE Press， 2020： 4537-4544.
[20]	Ultralytics. YOLOv8 2023 ［EB/OL］. .
[21]	邓立暖，尧新峰. 基于NSST的红外与可见光图像融合算法［J］. 电子学报，2017， 45（12）： 2965-2970.
	DENG L Y， YAO X F， et al ［J］. Infrared and visible image fusion algorithm based on NSST. Journal of Electronics， 2017， 45（12）： 2965-2970 （in Chinese）
[22]	DAI M， ZHENG E， FENG Z， et al. Vision-based UAV self-positioning in low-altitude urban environments［J］. IEEE Transactions on Image Processing， 2023， 33： 493-508.
[23]	HU J， SHEN L， SUN G. Squeeze-and-excitation networks［C］∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2018： 7132-7141.
[24]	ROY A G， NAVAB N， WACHINGER C. Concurrent spatial and channel ‘squeeze & excitation’in fully convolutional networks［C］∥Medical Image Computing and Computer Assisted Intervention-MICCAI 2018： 21st International Conference， 2018： 421-429.
[25]	CAO Y， XU J， LIN S， et al. Gcnet： Non-local networks meet squeeze-excitation networks and beyond［C］∥Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops. Piscataway： IEEE Press， 2019.
[26]	ZHANG P， RAN H， JIA C， et al. A lightweight propagation path aggregating network with neural topic model for rumor detection［J］. Neurocomputing， 2021， 458： 468-477.
[27]	ZHANG X， ZENG H， GUO S， et al. Efficient long-range attention network for image super-resolution［C］∥European Conference on Computer Vision. Cham： Springer Nature Switzerland， 2022： 649-667.
[28]	HE K， ZHANG X， REN S， et al. Deep residual learning for image recognition［C］∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2016： 770-778.
[29]	LIU X， PENG H， ZHENG N， et al. Efficientvit： Memory efficient vision transformer with cascaded group attention［C］∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2023： 14420-14430.
[30]	LIU Z， LIN Y， CAO Y， et al. Swin transformer： Hierarchical vision transformer using shifted windows［C］∥Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE Press， 2021： 10012-10022.
[31]	MEHTA S， RASTEGARI M. Mobilevit： Light-weight， general-purpose， and mobile-friendly vision transformer［DB/OL］. arXiv preprint：2110.02178， 2021.
[32]	CHEN J， KAO S， HE H， et al. Run， don’t walk： Chasing higher FLOPS for faster neural networks［C］∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE Press， 2023： 12021-12031.

E-mail：hkxb@buaa.edu.cn

关于我们

期刊社服务

专业学科

封面文章

友情链接

主管单位：中国科学技术协会主办单位：中国航空学会北京航空航天大学

超参数	数值
批大小（Batch size）	32
初始学习率（Learn rate）	0.01
迭代次数（Epoch）	300
优化器	随机梯度下降（SGD）
Mosic	False

模型	PFIS	TACM	MSFM	mAP50/%	mAP50-95/%	GFLOPs	Size/MB	Inference/ms
Yolov8n-seg				83.8	62.4	12.6	6.75	2.3
+PFIS	√			85.3	65.5	12.9	6.84	2.3
+TACM		√		86.1	66.7	14.7	8.52	2.5
+MSFM			√	85.0	64.6	13.3	7.99	2.1
+PFIS，TACM	√	√		86.9	68.7	14.9	8.64	2.6
DGISM	√	√	√	88.8	69.1	15.4	10.34	2.3

卷积模块	mAP50/%	mAP50-95/%	Infenence/ms
GroupConv	81.3	62.6	2.2
AtrousConv	83.5	62.7	2.5
TransposedConv	84.1	63.7	2.4
AsymConv	85.5	64.8	2.3
DWConv	85.3	65.5	2.1

注意力	mAP50/%	mAP50-95/%	GFLOPs	Inference/ms
SE	83.9	63.1	13.2	2.2
scSE	81.0	59.4	13.0	1.9
GCNet	80.5	60.1	12.9	2.1
PPA	85.8	66.2	19.3	4.4
LFE	86.0	67.6	14.6	3.3
TACM	86.9	68.7	14.9	2.6

特征	C+H	C+W	H+W	mAP50/%	mAP50-95/%	Inference/ms
SFIM	√			82.4	61.8	2.6
SFIM		√		85.9	65.9	2.3
SFIM			√	86.2	64.8	2.8
DFIM	√	√		85.8	66.8	2.6
DFIM	√		√	85.5	68.5	2.8
DFIM		√	√	84.6	63.4	2.7
TFIM	√	√	√	86.9	68.7	2.6

GPS拒止环境巡飞器双源地理位置知觉实例分割

Instance segmentation of dual-source geolocation perception for GPS reject environmental aerial vehicle

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 15

参考文献 32

相关文章 2

编辑推荐

Metrics

本文评价

模型	mAP50/%	mAP50-95/%	GFLOPs	Size/MB	Inference/ms
Yolov5n-seg	81.2	61.3	7.1	4.12	7.5
Yolov5s-seg	80.4	60.3	26.4	14.8	11.0
Yolov8n-seg	83.8	62.4	12.6	6.75	2.3
Yolov8s-seg	83.9	66.0	42.8	22.8	4.4
Yolov8m-seg	83.0	64.4	110.5	54.9	6.3
Yolov9c-seg	86.3	69.0	158.5	53.7	8.5
Yolov11n-seg	85.6	66.8	10.5	6.39	2.0
Yolov11s-seg	85.4	65.3	35.7	21.4	2.9
Yolov12n-seg	86.1	63.8	10.4	5.87	2.2
Yolov12s-seg	85.4	63.5	35.4	20.4	4.7
DGISM	88.8	69.1	15.4	10.34	2.3

2292-H1	Label	DGISM/10^-9	Label	Yolov8n-seg/10^-9
定位精度	DGISM 95.6%		Yolov8n-seg 76.1%
TOP1	2292-H1	0.235	2297-H3	2.878
TOP2	2292-H2	0.235	2297-H1	2.878
TOP3	2292-H3	0.235	2297-H2	2.878
TOP4	2304-H3	0.311	2292-H3	2.935
TOP5	2304-H1	0.311	2292-H2	2.935

[1]	万开方, 吴志林, 武韫晖, 强皓植, 吴艺博, 李波. 拒止环境下基于深度强化学习的多无人机协同定位[J]. 航空学报, 2025, 46(8): 331024-331024.
[2]	姚俞成, 李旭, 徐启敏, 孔栋. 基于特征增强和校准的航拍车辆实例分割方法[J]. 航空学报, 2023, 44(24): 328397-328397.

主干网络	mAP50/%	mAP50-95/%	GFLOPs	Size/MB	Inference/ms
Yolov8n-seg	83.8	62.4	12.6	6.75	2.3
Resnet50	82.3	64.3	80.2	55.6	7.2
EfficientViT	77.8	50.7	56.7	10.8	4.5
SwinTransformer	73.1	50.2	449.3	60.2	12.2
MobileViT	70.0	51.1	13.8	7.0	5.1
FasterNet	75.9	51.8	57.9	11.3	3.3
Yolov8n-seg+TACM，MSFM	86.6	67.2	15.2	9.75	2.5