A statistical method for line segment detection

Xu, Zezhong; Shin, Bok-Suk; Klette, Reinhard

doi:10.1016/j.cviu.2015.05.008

Cited by 23 publications

(10 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given its simplicity and effectiveness, subsequent line-detection work followed this approach [11,20,49], by focusing on analyzing peaks in Hough space. To overcome the sensitivity to noise, previous work proposed statistical analysis of Hough space [50], and segment-set selection based on hypothesis testing [45]. Similarly, a probabilistic Hough transform for line detection, followed by Markov Chain modelling of candidate lines is proposed in [1], while [26] creates a progressive probabilistic Hough transform, which is both faster and more robust to noise.…”

Section: Related Workmentioning

confidence: 99%

Deep Hough-Transform Line Priors

Lin

Pintea

Gemert

2020

Preprint

View full text Add to dashboard Cite

Classical work on line segment detection is knowledge-based; it uses carefully designed geometric priors using either image gradients, pixel groupings, or Hough transform variants. Instead, current deep learning methods do away with all prior knowledge and replace priors by training deep networks on large manually annotated datasets. Here, we reduce the dependency on labeled data by building on the classic knowledge-based priors while using deep networks to learn features. We add line priors through a trainable Hough transform block into a deep network. Hough transform provides the prior knowledge about global line parameterizations, while the convolutional layers can learn the local gradient-like line features. On the Wireframe (ShanghaiTech) and York Urban datasets we show that adding prior knowledge improves data efficiency as line priors no longer need to be learned from data.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep Hough-Transform Line Priors

Lin

Pintea

Gemert

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…A number of methods scan the detected lines in the image space looking for a maximal chain of connected or nearly-connected edges [19], [20]. Others have attempted to identify the endpoints of each line segment by analyzing the exact shape of a characteristic 'butterfly' pattern around the associated peak in the Hough map [21], [22], [23], [24], [25]. One major limitation of this approach is that only one segment can be found per line, whereas in built environments it is quite common to find multiple co-linear segments.…”

Section: The Hough Approachmentioning

confidence: 99%

MCMLSD: A Probabilistic Algorithm and Evaluation Framework for Line Segment Detection

Elder¹,

Almazàn²,

Qian³

et al. 2020

Preprint

View full text Add to dashboard Cite

Traditional approaches to line segment detection typically involve perceptual grouping in the image domain and/or global accumulation in the Hough domain. Here we propose a probabilistic algorithm that merges the advantages of both approaches. In a first stage lines are detected using a global probabilistic Hough approach. In the second stage each detected line is analyzed in the image domain to localize the line segments that generated the peak in the Hough map. By limiting search to a line, the distribution of segments over the sequence of points on the line can be modeled as a Markov chain, and a probabilistically optimal labelling can be computed exactly using a standard dynamic programming algorithm, in linear time. The Markov assumption also leads to an intuitive ranking method that uses the local marginal posterior probabilities to estimate the expected number of correctly labelled points on a segment. To assess the resulting Markov Chain Marginal Line Segment Detector (MCMLSD) we develop and apply a novel quantitative evaluation methodology that controls for under-and over-segmentation. Evaluation on the YorkUrbanDB and Wireframe datasets shows that the proposed MCMLSD method outperforms prior traditional approaches, as well as more recent deep learning methods.

show abstract

“…Line segment detection has attracted a lot of research work. Classical approaches [1,3,11,18,37,40] rely on low-level information, so are susceptible to external conditions. Recently, Wireframe [17] first adopts two independent networks to predict line and junction heatmaps parallelly, then combine junctions and lines to produce line segments.…”

Section: Related Workmentioning

confidence: 99%

“…The detected line segments further benefit numerous computer vision tasks, ranging from stereo matching [45] and 3D reconstruction [7,9,16,32,46] to image stitching [38] and segmentation [2,5]. Traditional techniques [1,3,11,18,37,40] based on hand-crafted features are vulnerable to textureless regions, repetitive textures, illumination variations, occlusions, etc. More recent deep learning approaches [17,41,47,49] attempt to explore semantic meanings of line segments to mitigate the problems.…”

Section: Introductionmentioning

confidence: 99%

LGNN: A Context-aware Line Segment Detector

Meng,

Zhang,

et al. 2020

Preprint

View full text Add to dashboard Cite

We present a novel real-time line segment detection scheme called Line Graph Neural Network (LGNN). Existing approaches require a computationally expensive verification or postprocessing step. Our LGNN employs a deep convolutional neural network (DCNN) for proposing line segment directly, with a graph neural network (GNN) module for reasoning their connectivities. Specifically, LGNN exploits a new quadruplet representation for each line segment where the GNN module takes the predicted candidates as vertexes and constructs a sparse graph to enforce structural context. Compared with the state-of-the-art, LGNN achieves near real-time performance without compromising accuracy. LGNN further enables time-sensitive 3D applications. When a 3D point cloud is accessible, we present a multi-modal line segment classification technique for extracting a 3D wireframe of the environment robustly and efficiently.

show abstract

A statistical method for line segment detection

Cited by 23 publications

References 38 publications

Deep Hough-Transform Line Priors

Deep Hough-Transform Line Priors

MCMLSD: A Probabilistic Algorithm and Evaluation Framework for Line Segment Detection

LGNN: A Context-aware Line Segment Detector

Contact Info

Product

Resources

About