An overview of edge and object contour detection

Yang, Daipeng; Peng, Bo; Al‐Huda, Zaid; Malik, Asad; Zhai, Donghai

doi:10.1016/j.neucom.2022.02.079

Cited by 44 publications

(21 citation statements)

References 162 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Currently, edge-detection techniques are widely used in many fields. These include edge-feature extraction for medical images, license-plate recognition, and face-detection techniques [ 23 , 24 ].…”

Section: Methodsmentioning

confidence: 99%

Extraction of Step-Feature Lines in Open-Pit Mines Based on UAV Point-Cloud Data

Mao

Wang

et al. 2022

Sensors

View full text Add to dashboard Cite

Step-feature lines are one of the important geometrical elements for drawing the status quo maps of open-pit mines, and the efficient and accurate automatic extraction and updating of step-feature lines is of great significance for open-pit-mine stripping planning and analysis. In this study, an automatic extraction method of step-feature lines in an open-pit mine based on unmanned-aerial-vehicle (UAV) point-cloud data is proposed. The method is mainly used to solve the key problems, such as low accuracy, local-feature-line loss, and the discontinuity of the step-feature-line extraction method. The method first performs the regular raster resampling of the open-pit-mine cloud based on the MLS algorithm, then extracts the step-feature point set by detecting the elevation-gradient change in the resampled point cloud, further traces the step-feature control nodes by the seed-growth tracking algorithm, and finally generates smooth step-feature lines by fitting the space curve to the step-feature control nodes. The results show that the method effectively improves the accuracy of step-feature-line extraction and solves the problems of local-feature-line loss and discontinuity.

show abstract

Section: Methodsmentioning

confidence: 99%

Extraction of Step-Feature Lines in Open-Pit Mines Based on UAV Point-Cloud Data

Mao

Wang

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Roberts' method has the disadvantage of being severely affected by noise [21]. In addition, roberts operators give a weak response to the edges, unless the edges are very sharp [22].…”

Section: A Methods Robertsmentioning

confidence: 99%

Image Edge Detection Capture Zoom for Facial Recognition Using Gradient Operators

Hadi

Sumijan

2022

komtekinfo

View full text Add to dashboard Cite

Imagery on edge detection is a process that will display the edges of an image. Basically, edge detection is one of the techniques for the analysis of image quality in the spatial domain and is also one of the initial process in digital image processing. Edge detection serves to detect the border line of an object contained in the image. This study aims to identify and recognize face pattern objects in the capture zoom image. To perform face identification begins with collecting image data, image processing, image edge detection, thinning of the image, and identification process using the template matching method. The method used in edge detection uses 3 methods, namely Sobel, Roberts and Prewitt which are gradient operators to detect edges in facial images. The dataset used is image capture zoom. The trial was carried out in two stages, namely the identification of the face shape and the identification of the edges of the face. The conclusion of the study is that the Roberts operator is the operator that finds the least edge patterns in facial images than the other two operators, namely Prewitt and Sobel. Meanwhile, the Sobel operator produces edge patterns that are better in quality and quantity than using the Roberts and Prewitt operators.

show abstract

“…Since the inception of Sobel edge detector [12], back in the 60s of the last century, a large number of approaches have been proposed for tackling this challenging task. For an in depth review, the following edge detection surveys can be considered [13]- [17]. Edge detectors can be broadly categorized into four groups: i) driven by low-level features; ii) brain-inspired; iii) classical learning-based; and iv) deep learning as stated in [2].…”

Section: Related Workmentioning

confidence: 99%

LDC: Lightweight Dense CNN for Edge Detection

2022

View full text Add to dashboard Cite

This paper presents a Lightweight Dense Convolutional (LDC) neural network for edge detection. The proposed model is an adaptation of two state-of-the-art approaches, but it requires less than 4% of parameters in comparison with these approaches. The proposed architecture generates thin edge maps and reaches the highest score (i.e., ODS) when compared with lightweight models (models with less than 1 million parameters), and reaches a similar performance when compare with heavy architectures (models with about 35 million parameters). Both quantitative and qualitative results and comparisons with stateof-the-art models, using different edge detection datasets, are provided. The proposed LDC does not use pre-trained weights and requires straightforward hyper-parameter settings. The source code is released at https://github.com/xavysp/LDC.

show abstract

An overview of edge and object contour detection

Cited by 44 publications

References 162 publications

Extraction of Step-Feature Lines in Open-Pit Mines Based on UAV Point-Cloud Data

Extraction of Step-Feature Lines in Open-Pit Mines Based on UAV Point-Cloud Data

Image Edge Detection Capture Zoom for Facial Recognition Using Gradient Operators

LDC: Lightweight Dense CNN for Edge Detection

Contact Info

Product

Resources

About