2013
DOI: 10.1109/tip.2013.2274388
|View full text |Cite
|
Sign up to set email alerts
|

Multi-Class Multi-Scale Series Contextual Model for Image Segmentation

Abstract: Contextual information has been widely used as a rich source of information to segment multiple objects in an image. A contextual model uses the relationships between the objects in a scene to facilitate object detection and segmentation. Using contextual information from different objects in an effective way for object segmentation, however, remains a difficult problem. In this paper, we introduce a novel framework, called multiclass multiscale (MCMS) series contextual model, which uses contextual information… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
21
0

Year Published

2014
2014
2021
2021

Publication Types

Select...
4
2

Relationship

0
6

Authors

Journals

citations
Cited by 30 publications
(21 citation statements)
references
References 37 publications
0
21
0
Order By: Relevance
“…A membrane detection map ℬ = { b i } is a probability image with each pixel intensity b i ∈ [0, 1] as shown in Figure 1(b). Such membrane detection maps are commonly obtained using machine learning algorithms, such as (Jain et al, 2007; Laptev et al, 2012; Ciresan et al, 2012; Seyedhosseini et al, 2013). We can obtain a segmentation of the cells in the image by simply thresholding the probability map.…”
Section: Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…A membrane detection map ℬ = { b i } is a probability image with each pixel intensity b i ∈ [0, 1] as shown in Figure 1(b). Such membrane detection maps are commonly obtained using machine learning algorithms, such as (Jain et al, 2007; Laptev et al, 2012; Ciresan et al, 2012; Seyedhosseini et al, 2013). We can obtain a segmentation of the cells in the image by simply thresholding the probability map.…”
Section: Methodsmentioning
confidence: 99%
“…The boundary classifier takes 88 non-local features computed from each pair of merging regions, including region geometry, image intensity, and texture statistics from both original EM images and membrane detection probability maps, and merging saliency information (see Appendix B for a summary of features). One advantage of our features over the local features commonly used by pixel classifiers (Laptev et al, 2012; Seyedhosseini et al, 2013) is that our features are extracted from regions instead of pixels and thus can be more informative. For instance, we use geometric features to incorporate region shape information for the classification procedure, which is not feasible for pixel classifiers.…”
Section: Methodsmentioning
confidence: 99%
“…Second row: Horse segmentation (Weiz-mann dataset [17]. Third row: Membrane detection (mouse neuropil dataset [9]). Fourth row: Edge detection (Berkeley dataset [19]).…”
Section: Figmentioning
confidence: 99%
“…However, they have a drawback that they do not obtain contextual information at multiple scales. Multi-scale processing of images has been proven critical in many computer vision tasks [8], [9]. OWT-UCM [10] takes advantage of processing the input image at multiple scales through a hierarchy.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation