Improved multi-scale edge detection method based on HED

ZHANG Xiao-wen; REN Yong-feng

Chinese

您当前的位置：

首页 >

文章列表页 >

Improved multi-scale edge detection method based on HED

更新时间：2026-09-04

- Improved multi-scale edge detection method based on HED
- Vol. 38, Issue 6, Pages: 1-6(2021)
- 作者机构：
  
  中北大学机构电子测试技术国家重点实验室,山西,太原,030051
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published：2021
- 稿件说明：
移动端阅览
张晓闻, 任勇峰. 基于改进HED的多尺度边缘检测方法[J]. 微电子学与计算机, 2021, 38(6): 1-6, 12.

ZHANG Xiao-wen, REN Yong-feng. Improved multi-scale edge detection method based on HED[J]. 2021, 38(6): 1-6.
张晓闻, 任勇峰. 基于改进HED的多尺度边缘检测方法[J]. 微电子学与计算机, 2021, 38(6): 1-6, 12. DOI：

ZHANG Xiao-wen, REN Yong-feng. Improved multi-scale edge detection method based on HED[J]. 2021, 38(6): 1-6. DOI：

摘要

针对基于深度学习的多尺度边缘检测不可避免出现自适应性低，参数增加，计算量大，检测边缘不连续等问题，本文提出一种基于改进整体嵌套的多尺度边缘检测方法.该方法将多尺度检测与弱监督模型相结合，解决参数多计算量大的问题.为了充分利用卷积强大的特征表达能力，在整体嵌套边缘检测的基础上，提出了一种多尺度下深度学习结构，一个相互独立的多网络多尺度结构，由不同深度和输出的多个网络组合.同时引用整体嵌套的权重混合层，权重混合层将所有的弱监督预测结果连接到一起，并在训练的过程中学习混合权重。将所提方法在数据集BSDS500上通过评价指标进行性能评价.实验结果表明，所提方法能够在数据集BSDS500上达到很好的性能.

Abstract

Aiming at the problems of deep learning-based multi-scale edge detection inevitably low adaptability

increased parameters

large calculations

and discontinuous detection edges

this paper proposes a multi-scale edge detection method based on improved overall nesting. The method combines multi-scale detection with weak supervision model to solve the problem of large amount of parameter calculation. In order to make full use of the powerful feature expression ability of convolution

based on the whole nested edge detection

a multi-scale deep learning structure is proposed

which is an independent multi-network multi-scale structure

which is composed of multiple networks with different depths and outputs.At the same time

the whole nested weight mixing layer is referenced. The weight mixing layer connects all the weak supervised prediction results together and learns the mixed weight in the training process. The performance of the proposed method is evaluated by the evaluation index on the data set BSDS500. Experimental results show that the proposed method can achieve good performance on BSDS500.

关键词

Keywords

references

WITKIN A P. Scale-space filtering[C]//Proceedings of the Eighth International Joint Conference on Artificial Intelligence. San Francisco: Morgan Kaufmann Publishers Inc., 1983: 1019-1022.

KHOREVA A, BENENSON R, OMRAN M, et al. Weakly supervised object boundaries[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 183-192. DOI: 10.1109/CVPR.2016.27.

ARBELAEZ P, MAIRE M, Fowlkes C, et al. Contour detection and hierarchical image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(5): 898-916. DOI: 10.1109/TPAMI.2010.161.

ZHOU Z H. A brief introduction to weakly supervised learning[J]. National Science Review, 2018, 5(1): 44-53. DOI: 10.1093/nsr/nwx106.

XIE S N, TU Z W. Holistically-nested edge detection[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Santiago: IEEE, 2015: 1395-1403. DOI: 10.1109/ICCV.2015.164.

PATI Y C, REZAⅡFAR R, KRISHNA P S. Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition[C]//Proceedings of the 27th Asilomar Conference on Signals, Systems and Computers. Pacific Grove: IEEE, 1993: 1412-1416. DOI: 10.1109/ACSSC.1993.342465.

ZHANG Z Z, XING F Y, SHI X S, et al. SemiContour: a semi-supervised learning approach for contour detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 251-259. DOI: 10.1109/CVPR.2016.34.

OBUKHOV A, GEORGOULIS S, DAI D X, et al. Gated CRF loss for weakly supervised semantic image segmentation[J]//Neur IPS, Vancouver, 2019.

DOLLAR P, ZITNICK C L. Fast edge detection using structured forests[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(8): 1558-1570. DOI: 10.1109/TPAMI.2014.2377715.

WANG L Y, SHEN Y, LIU H D, et al. An accurate and efficient multi-category edge detection method[J]. Cognitive Systems Research, 2019, 58: 160-172. DOI: 10.1016/j.cogsys.2019.06.002.

MARR D, HILDRETH E. Theory of edge detection[J]. Proceedings of the Royal Society of London. Series B, Biological Sciences, 1980, 207(1167): 187-217. DOI: 10.1098/rspb.1980.0020.

REN X F, BO L F. Discriminatively trained sparse code gradients for contour detection[C]//Proceedings of the 25th International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2012, 1: 584-592.

ROSENFELD A, THURSTON M. Edge and curve detection for visual scene analysis[J]. IEEE Transactions on Computers, 1971, C-20(5): 562-569. DOI: 10.1109/T-C.1971.223290.

AHARON M, ELAD M, BRUCKSTEIN A. K-SVD: an algorithm for designing overcomplete dictionaries for sparse representation[J]. IEEE Transactions on Signal Processing, 2006, 54(11): 4311-4322. DOI: 10.1109/TSP.2006.881199.

LI Y, PALURI M, REHG J M, et al. Unsupervised learning of edges[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 1619-1627. DOI: 10.1109/CVPR.2016.179.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Aerial small object detection algorithm based on improved RT-DETR

Research on assembly action recognition based on MSS-YOLO-MPHands fusion architecture

A high-efficiency charge pump for ultra-low voltage boost systems

Steganography model guided by attention mechanism and improved dense convolution

Implementation and verification of a dual-chain power gating strategy for low power design

Related Author

No data

Related Institution

No data

AI问答

Postal code：100079
Tel：（010）53879206 Email：tmw@bjxintong.com.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09082226号-64 京公网安备11010602201714号
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰