采用脉冲耦合神经网络的改进显著性区域提取方法

贾松敏; 徐涛; 董政胤; 李秀智

doi:10.3788/OPE.20152303.0819

您当前的位置：

首页 >

文章列表页 >

采用脉冲耦合神经网络的改进显著性区域提取方法

信息科学 | 更新时间：2020-08-13

- 采用脉冲耦合神经网络的改进显著性区域提取方法
- Improved salience region extraction algorithm with PCNN
- 光学精密工程 2015年23卷第3期页码：819-826
- 作者机构：
  
  1. 河南科技学院机电学院,河南新乡,中国,453003
  2. 北京工业大学电子信息与控制工程学院北京,100124
- 作者简介：
- 基金信息：
  
  国家自然科学基金资助项目(No.61175087,No.61105033)()
- DOI：10.3788/OPE.20152303.0819
  中图分类号： TP391.4
- 收稿日期：2014-10-30，
  
  修回日期：2014-12-05，
  
  纸质出版日期：2015-03-25
- 稿件说明：
移动端阅览
贾松敏, 徐涛, 董政胤等. 采用脉冲耦合神经网络的改进显著性区域提取方法[J]. 光学精密工程, 2015,23(3): 819-826

JIA Song-min, XU Tao, DONG Zheng-yin etc. Improved salience region extraction algorithm with PCNN[J]. Editorial Office of Optics and Precision Engineering, 2015,23(3): 819-826
贾松敏, 徐涛, 董政胤等. 采用脉冲耦合神经网络的改进显著性区域提取方法[J]. 光学精密工程, 2015,23(3): 819-826 DOI： 10.3788/OPE.20152303.0819.

JIA Song-min, XU Tao, DONG Zheng-yin etc. Improved salience region extraction algorithm with PCNN[J]. Editorial Office of Optics and Precision Engineering, 2015,23(3): 819-826 DOI： 10.3788/OPE.20152303.0819.

摘要

由于仅考虑颜色等视觉对比信息的视觉显著性提取模型不符合人眼生物学过程

本文提出了一种基于混合模型的改进显著性区域提取(ISRE)方法。该混合模型由显著性滤波算法和改进脉冲耦合神经网络(PCNN)算法构成。首先

利用显著性滤波器算法获得原图像的初始显著性图(OSM)和亮度特征图(IFM)

用IFM作为PCNN的输入神经元;然后

进一步对PCNN点火脉冲输入进行改进

即对PCNN内部神经元与OSM的二值化显著性图进行点乘

确定最终点火脉冲输入

以获得更加准确的点火范围;最后

通过改进后的PCNN多次迭代

完成显著性二值化区域提取。基于1 000 张标准图像数据库进行的实验结果显示:在视觉效果和客观定量数据比对两方面

本算法均优于现有的5种显著性提取方法

平均查准率为0.891

平均召回率为0.808

综合指标

值为0.870。在真实环境实验中

所提算法获得了精确的提取效果

进一步验证了本算法具有较高的准确性和执行效率。

Abstract

The visual salience extraction model only considers visual contrasting information and it does not conform to the biology process of human eyes. Therefroe

a hybrid model based on Improved Salient Region Extraction (ISRE) algorithm was proposed in this paper. This hybrid model consists of a salience filtering algorithm and an improved Pulse Coupled Neural Network (PCNN) algorithm. Firstly

the salience filtering algorithm was used to get Original Salience Map (OSM) and Intensity Feature Map (IFM) was used as the input neuron of PCNN. Then

the PCNN ignition pulse input was further improved as follows: the point multiplication algorithm was taken between the PCNN internal neuron and the binarization salience image of OSM to determine the final ignition pulse input and to make the ignition range more exact. Finally

the salience binarization region was extracted by the improved PCNN multiply iteration. Based on ASD standard data base

some experiments on 1 000 images were performed. The experimental results show that the proposed algorithm is superior to the five existing salience extraction algorithms uniformly in visual effect and objective quantitative data comparison. The results display that the precision ratio

recall ratio

and the overall

-measure of the proposed extraction algorithm are 0.891

0.808

and 0.870

respectively. In a real context experiment

the proposed algorithm gets more accurate extraction effect

which verifies that the proposed algorithm has higher accuracy and execution efficiency.

关键词

Keywords

references

寿天德. 视觉信息处理的脑机制 [M]. 合肥:中国科学技术大学出版社, 2010. SHOU T D. Brain Mechanisms of Visual Information Processing [M]. Hefei:Press of University of Science and Technology of China, 2010. (in Chinese)

KOCH C, ULLMAN S. Shifts in selective visual attention:towards the underlying neural circuitry [J]. Human neurobiology, 1985, 4(4):219-227.

LIU ZH, ZHANG X, LUO SH H, et al. . Superpixel-based spatiotemporal saliency detection [J]. IEEE Trans. Circuits Syst. Video Technol., 2014, 24(9):1522-1540.

王丽佳, 贾松敏, 王爽, 等. 采用改进Mean Shift算法的移动机器人行人跟踪 [J]. 光学精密工程, 2013, 21(9):2364-2370. WANG L J, JIA S M, WANG SH, et al. . Person tracking of mobile robot using improved Mean Shift [J]. Opt. Precision Eng. , 2013, 21(9):2364-2370. (in Chinese)

ZHANG L, SHEN Y, LI H, et al. . VSI:A visual saliency-induced index for perceptual image quality assessment [J]. IEEE Trans. Image Process. , 2014, 23(10):4270-4281.

MA L, LI S N, NGAN K N. Visual horizontal effect for image quality assessment [J]. IEEE Signal Process. Lett. , 2010, 17(7):627-630.

ITTI L, KOCH C, NIEBUR E. A model of saliency-based visual attention for rapid scene analysis [J]. IEEE TPAMI, 1998, 20(11):1254-1259.

HOU X, ZHANG L. Saliency detection:A spectral residual approach [C]. 2007 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Minneapolis, United states, CVPR, 2007:1-8.

STAS G, LIHI Z, AYELLET T. Context-aware saliency detection [C]. 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Francisco, United states, CVPR, 2010:2376-2383.

CHENG M, ZHANG G, MITRA N J, et al. . Global contrast based salient region detection [C]. 2011 IEEE Conference on Computer Vision and Pattern Recognition, Colorado Springs, United states, CVPR, 2011:409-416.

RADHAKRISHNA A, SHEILA H, FRANCISCO E, et al. . Frequency-tuned salient region detection [C]. 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, Miami, United states, CVPR Workshops, 2009:1597-1604.

PERAZZI F, KRAHENBUL P, PRITCH Y, et al. . Saliency filters:contrast based filtering for salient region detection [C]. 2012 IEEE Conference on Computer Vision and Pattern Recognition, Providence, United states, CVPR, 2012:733-740.

张颖颖, 张帅, 张萍, 等. 融合对比度和分布性的图像显著性区域检测 [J]. 光学精密工程, 2014, 22(4):1012-1019. ZHANG Y Y, ZHANG SH, ZHANG P, et al. . Detection of salient maps by fusion of contrast and distribution [J]. Opt. Precision Eng. , 2014, 22(4):1012-1019. (in Chinese)

RICHARD J, DANNY C. Deep salience:visual salience modeling via deep belief propagation [C]. Twenty-Eighth AAAI Conference on Artificial Intelligence, Québec City, Canada, AAAI, 2014:2773-2779.

HODGKIN A L, HUXLEY A F. A quantitative description of membrane current and its application to conduction and excitation in nerve [J]. The Journal of physiology, 1952, 117(4):500-544.

ECKHORN R, REITNOECK H J, ARNDT M, et al.. Feature linking via synchronization among distributed assemblies:Simulations of results from cat visual cortex [J]. Neural Computation, 1990, 2(3):293-307.

THOMAS L, JASON M K. Image Processing Using Pulse-Coupled Neural Networks [M]. New York:Springer-Verlag, 2013.

CHEN Y, PARK S K, MA Y, et al.. A new automatic parameter setting method of a simplied PCNN for image segmentation [J]. IEEE Trans. Neural Netw., 2011, 22(6):880-892.

郑欣, 彭真明. 基于活跃度的脉冲耦合神经网络图像分割 [J]. 光学精密工程, 2013, 21(3):821-827. ZHENG X, PENG ZH M. Image segmentation based on activity degree with pulse coupled neural networks [J]. Opt. Precision Eng. , 2013, 21(3):821-827. (in Chinese)

LIU S, HE D, LIANG X.. An improved hybrid model for automatic salient region detection [J]. IEEE Signal Processing Letters, 2012, 19(4):207-210.

王蒙军, 阳路, 王霞, 等. 用简化脉冲耦合神经网络实现交通标志图像的类Euclidean距离变化类内特征提取 [J]. 光学精密工程, 2012, 20(12):2751-2758. WANG M J, YANG L, WANG X, et al. . Realization of within-class feature extracting based on Euclidean-like distance transform for traffic signs using simplied PCNN [J]. Opt. Precision Eng. , 2012, 20(12):2751-2758. (in Chinese)

MARTIN D, FOWLKES C, TAL D, et al.. A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics [C]. 8th IEEE International Conference on Computer Vision, Vancouver, United states, ICCV, 2001, 2:416-423.

浏览量

243

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于分数阶微分的高光谱图像特征提取与分类

面向高光谱影像场景分类的轻量化深度全局-局部知识蒸馏网络

多分支无锚框网络密集行人检测算法

SENet生成对抗网络在图像语义描述中的应用

面向高光谱图像分类的半监督双流网络