一种基于定位和非对称补偿的伪装目标分割方法

doi:10.12305/j.issn.1001-506X.2022.09.02

摘要/Abstract

摘要：

伪装是欺骗观察者感知系统的一种手段, 善于伪装的个体在纹理特征上与背景具有高度的相似性。为解决前景与背景因相似而导致的像素归属歧义, 提出一种基于定位和补偿网络(locating and compensation network, LCNet)的伪装目标分割网络。该方法效仿了捕食者从搜索→确立→聚焦的寻猎过程, 涵盖双主干网的强感知提取、定位模块的双注意力以及级联的非对称补偿模块的细化像素模糊。实验表明, 在4种评价指标下, LCNet在3个具有挑战的伪装数据集上都显著优于现有的6种最新模型, 具有较高分割性能。

关键词: 纹理伪装目标, 非对称注意力补偿, 双注意力定位, 双主干网

Abstract:

Camouflage is the instinct to deceive the perception of the observer, which presents the high similarity of the textural characteristics with the surroundings. In order to address the ambiguous regions generated by the similarity between background and foreground, this paper proposes a camouflaged object segmentation network based on locating and compensation network (LCNet). Inspired by predator process: Search→Establish→Focusing, the paradigm of the proposed method is achieved via dual-backbone with the strong sense of knowledge extraction, locating module with double attention, and cascading asymmetric compensation module with pixel refining. Experimental results have shown that the performances of LCNet are superior than six state-of-the-art models at the three major challenging camouflaged datasets in terms of four metrics, and the effectiveness of LCNet is demonstrated.

Key words: textural camouflaged object, asymmetric attention compensation, double attention locating, dual-backbone

中图分类号:

TP391.41

徐义飞, 李晓冬, 李新德. 一种基于定位和非对称补偿的伪装目标分割方法[J]. 系统工程与电子技术, 2022, 44(9): 2707-2715.

Yifei XU, Xiaodong LI, Xinde LI. A method of camouflaged object segmentation with locating and asymmetric compensation[J]. Systems Engineering and Electronics, 2022, 44(9): 2707-2715.

图/表 9

图1

图2

图3

图4

图5

表1

图6

表2

图7

参考文献 38

1	STEVENS M , MERILAITA S . Animal camouflage: current issues and new perspectives[J]. Philosophical Transaction of the Royal Society B: Biological Sciences, 2009, 364 (1516): 423- 427. doi: 10.1098/rstb.2008.0217
2	ZHENG Y W , ZHANG X W , WANG F , et al. Detection of people with camouflage pattern via dense deconvolution network[J]. IEEE Signal Processing Letters, 2018, 26 (1): 29- 33.
3	LE T N , NGUYEN T V , NIE Z , et al. Anabranch network for camouflaged object segmentation[J]. Computer Vision and Image Understanding, 2019, 184, 45- 56. doi: 10.1016/j.cviu.2019.04.006
4	FAN D P, JI G P, SUN G, et al. Camouflaged object detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 2777-2787.
5	LIU T , YUAN Z J , SUN J , et al. Learning to detect a salient object[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2010, 33 (2): 353- 367.
6	WANG L J, LU H C, WANG Y F, et al. Learning to detect salient objects with image-level supervision[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 136-145.
7	SHI J P , YAN Q , XU L , et al. Hierarchical image saliency detection on extended CSSD[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2015, 38 (4): 717- 729.
8	LI G B, YU Y Z. Visual saliency based on multiscale deep features[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 5455-5463.
9	MOVAHEDI V, ELDER J H. Design and perceptual validation of performance measures for salient object segmentation[C]//Proc. of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2010: 49-56.
10	TROSCIANKO T , BENTON C P , LOVELL P G , et al. Camouflage and visual perception[J]. Philosophical Transaction of the Royal Society B: Biological Sciences, 2009, 364 (1516): 449- 461. doi: 10.1098/rstb.2008.0218
11	周宇, 张慧, 冷沙. 战场上的"伪装者"——伪装技术对作战的影响[J]. 军事文摘, 2020, (7): 16- 20.
	ZHOU Y , ZHANG H , LENG S . "Pretenders" on the battlefield—the influence of camouflage technology on combat[J]. Military Digest, 2020, (7): 16- 20.
12	鲜晓东, 李克文. 基于颜色和纹理特征的伪装色矿工目标检测[J]. 计算机应用, 2013, 33 (2): 539- 542.
	XIAN X D , LI K W . Camouflage miner target detection based on color and texture features[J]. Computer Applications, 2013, 33 (2): 539- 542.
13	FAN D P, JI G P, ZHOU T, et al. Pranet: parallel reverse attention network for polyp segmentation[C]//Proc. of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2020: 263-273.
14	SENGOTTUVELAN P, WAHI A, SHANMUGAM A. Performance of decamouflaging through exploratory image analysis[C]//Proc. of the IEEE International Conference on Emerging Trends in Engineering and Technology, 2008: 6-10.
15	PAN Y X , CHEN Y W , FU Q , et al. Study on the camouflaged target detection method based on 3D convexity[J]. Modern Applied Science, 2011, 5 (4): 152- 157.
16	CHEN Q, WANG Y M, YANG T, et al. You only look one-level feature[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 13039-13048.
17	SUN P, ZHANG W, WANG H, et al. Deep RGB-D saliency detection with depth-sensitive attention and automatic multi-modal fusion[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 1407-1417.
18	BOUDIAF M, KERVADEC H, MASUD Z I, et al. Few-shot segmentation without meta-learning: a good transductive inference is all you need?[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 13979-13988.
19	LV Y Q, ZHANG J, DAI Y C, et al. Simultaneously localize, segment and rank the camouflaged objects[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 11591-11601.
20	HALL J R, CUTHILL I C, BADDELEY R, et al. Camouflage, detection and identification of moving targets[J]. Proceedings of the Royal Society B: Biological Sciences, 2013, 280(1758): 20130064.
21	金国栋, 薛远亮, 谭力宁, 等. 基于孪生神经网络的目标跟踪算法进展研究[J]. 系统工程与电子技术, 2022, 44 (6): 1805- 1822.
	JIN G D , XUE Y L , TAN L N , et al. Research on the progress of target tracking algorithm based on twin neural network[J]. Systems Engineering and Electronics, 2022, 44 (6): 1805- 1822.
22	曹旭, 邹焕新, 成飞, 等. 基于RHTC网络的飞机目标检测与精细识别[J]. 系统工程与电子技术, 2021, 43 (12): 3439- 3451. doi: 10.12305/j.issn.1001-506X.2021.12.04
	CAO X , ZOU H X , CHENG F , et al. Aircraft target detection and fine recognition based on RHTC network[J]. Systems Engineering and Electronics, 2021, 43 (12): 3439- 3451. doi: 10.12305/j.issn.1001-506X.2021.12.04
23	LIU Y D, WANG Y T, WANG S W, et al. Cbnet: a novel composite backbone network architecture for object detection[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2020: 11653-11660.
24	DING X H, ZHANG X Y, HAN J G, et al. Diverse branch block: building a convolution as an inception-like unit[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 10886-10895.
25	LEE Y W, PARK J. Centermask: real-time anchor-free instance segmentation[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 13906-13915.
26	WANG X, GIRSHICK R, GUPTA A, et al. Non-local neural networks[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 7794-7803.
27	DE BOER P T , KROESE D P , MANNOR S , et al. A tutorial on the cross-entropy method[J]. Annals of Operations Research, 2005, 134 (1): 19- 67. doi: 10.1007/s10479-005-5724-z
28	QIN X B, ZHANG Z C, HUANG C Y, et al. Basnet: boundary-aware salient object detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 7479-7489.
29	WANG Z, SIMONCELLI E P, BOVIK A C. Multi-scale structural similarity for image quality assessment[C]//Proc. of the IEEE 37th Asilomar Conference on Signals, Systems & Computers, 2003: 1398-1402.
30	FAN D P, GONG C, CAO Y, et al. Enhanced-alignment measure for binary foreground map evaluation[C]//Proc. of the 27th International Joint Conference on Artificial Intelligence, 2018: 698-703.
31	FAN D P, CHENG M M, LIU Y, et al. Structure-measure: a new way to evaluate foreground maps[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 4548-4557.
32	MARGOLIN R, ZELNIK-MANOR L, TAL A. How to evaluate foreground maps?[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 248-255.
33	PERAZZI F, KRÄHENBVHL P, PRITCH Y, et al. Saliency filters: contrast based filtering for salient region detection[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2012: 733-740.
34	ZHAO J X, LIU J J, FAN D P, et al. EGNet: edge guidance network for salient object detection[C]//Proc. of the IEEE/CVF International Conference on Computer Vision, 2019: 8779-8788.
35	WANG B, CHEN Q, ZHOU M, et al. Progressive feature polishing network for salient object detection[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2020: 12128-12135.
36	QIN X B , ZHANG Z C , HUANG C Y , et al. U2-net: going deeper with nested ustructure for salient object detection[J]. Pattern Recognition, 2020, 106, 107404. doi: 10.1016/j.patcog.2020.107404
37	PANG Y W, ZHAO X Q, ZHANG L H, et al. Multi-scale interactive network for salient object detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 9413-9422.
38	WEI J, WANG S H, HUANG Q M. F3Net: fusion, feedback and focus for salient object detection[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2020: 12321-12328.

方法	发布年份	COD10K-Test(2026张）				CAMO-Test(250张）				CPD1K-Test(200张）
方法	发布年份	E_ϕ↑	S_α↑	F_β^ω↑	M↓	E_ϕ↑	S_α↑	F_β^ω↑	M↓	E_ϕ↑	S_α↑	F_β^ω↑	M↓
EGNe^[34]	2019 ICCV	0.780	0.735	0.504	0.052	0.769	0.736	0.587	0.106	0.505	0.589	0.248	0.018
PFPN^[35]	2020 AAAI	0.779	0.745	0.484	0.062	0.778	0.736	0.572	0.116	0.791	0.786	0.434	0.016
U-2-net^[36]	2020 PR	0.737	0.708	0.482	0.064	0.645	0.634	0.447	0.129	0.904	0.881	0.749	0.005
MINet^[37]	2020 CVPR	0.837	0.767	0.600	0.048	0.802	0.747	0.635	0.095	0.881	0.851	0.704	0.005
F3Net^[38]	2020 AAAI	0.831	0.788	0.620	0.046	0.846	0.784	0.672	0.088	0.917	0.883	0.774	0.005
SINet^[4]	2020 CVPR	0.804	0.770	0.550	0.053	0.769	0.749	0.604	0.102	0.866	0.846	0.589	0.011
平均值		0.795	0.752	0.540	0.054	0.768	0.731	0.586	0.106	0.811	0.806	0.583	0.010
本文（Ⅰ）		0.873	0.801	0.660	0.039	0.859	0.785	0.697	0.083	0.950	0.889	0.794	0.004
本文（Ⅱ)		0.808	0.719	0.595	0.105	0.808	0.719	0.595	0.105	0.758	0.743	0.489	0.014

训练方案	消融对照			COD10K-test
训练方案	DB	LM双注意力	ACM补偿	E_ϕ↑	S_α↑	F_β^ω↑	M↓
方案Ⅰ	√	√	√	0.873	0.801	0.660	0.039
方案Ⅰ		√	√	0.859	0.757	0.623	0.043
方案Ⅰ	√		√	0.843	0.740	0.617	0.048
方案Ⅰ	√	√	√	0.856	0.753	0.633	0.053
方案Ⅰ	√	√		0.797	0.708	0.597	0.121
COD10K	√	√	√	0.808	0.719	0.595	0.105
CAMO	√	√	√	0.784	0.692	0.566	0.125
CPD1K	√	√	√	0.742	0.634	0.513	0.184

[1]	范加利, 田少兵, 黄葵, 朱兴动. 基于Faster R-CNN的航母舰面多尺度目标检测算法[J]. 系统工程与电子技术, 2022, 44(1): 40-46.
[2]	吕鑫, 慕晓冬, 张钧. 基于改进麻雀搜索算法的多阈值图像分割[J]. 系统工程与电子技术, 2021, 43(2): 318-327.
[3]	罗钧, 刘建强, 庞亚男. 基于邻域搜索JADE的二维Otsu多阈值图像分割[J]. 系统工程与电子技术, 2020, 42(10): 2164-2171.
[4]	花思齐, 赵伟, 刘建业. 基于改进滤波器和图像多尺度变换的背景抑制算法[J]. 系统工程与电子技术, 2020, 42(8): 1679-1684.
[5]	王晓龙, 刘海颖, 王景琪. 基于分层SLAM的空地多智能体协同导航[J]. 系统工程与电子技术, 2020, 42(1): 166-171.
[6]	康牧, 王宝树. 一种基于弹性模型的图像缩小算法[J]. Journal of Systems Engineering and Electronics, 2009, 31(9): 2208-2211.
[7]	陈洪波, 汤井田, 陈真诚. 同场景多幅图像斑纹噪声抑制方法[J]. Journal of Systems Engineering and Electronics, 2009, 31(5): 1051-1054.
[8]	陈尚锋, 肖山竹, 卢焕章. 图像序列弱小目标能量积累检测研究[J]. Journal of Systems Engineering and Electronics, 2009, 31(2): 288-291.
[9]	赵玉鑫, 刘光杰, 戴跃伟, 王执铨. 图像鲁棒哈希认证方案[J]. Journal of Systems Engineering and Electronics, 2009, 31(2): 292-295.