Synergism in low level vision

Christoudias, C. Mario; Georgescu, Bogdan; Meer, Peter

doi:10.1109/icpr.2002.1047421

Cited by 286 publications

(215 citation statements)

References 14 publications

Supporting

Mentioning

210

Contrasting

Unclassified

Order By: Relevance

“…First, the Berkeley image dataset does not have calibrated images and, consequently, we can not assure a good transformation from sRGB to CIE Luv. Second, because the size of L, u and v, is not the same and the method will require six parameters, instead of two, that is, − → σ L , − → σ u Figure 5 shows some results for the mean shift segmentation, corresponding to (h s , h r ) = { (7,15), (13,19), (17,23), (20,25), (25,30), (30, 35)}. These results point out the main advantage of RAD in favor of MS, namely, the capability of RAD to capture the DS of a histogram, whereas MS is ignorant to the physical processes underlying the structure of the DSs as Abd-Almageed and S. Davis explain in [10].…”

Section: Results and Performance Evaluationmentioning

confidence: 99%

“…Firstly, RAD is compared with Mean Shift (MS) [8], [22]. MS has been chosen because it is widely used, has a public available version, the EDISON one [23] and it has demonstrated its good performance [24]. Additionally, Mean Shift is a feature space analysis technique, as well as RAD, and yields a segmentation in a rather reasonable time, in opposition to other set of methods such as the Graph-Based approaches [25] , (with the exception of the efficient graph-based segmentation method introduced in [26]).…”

Section: Results and Performance Evaluationmentioning

confidence: 99%

“…Mean Shift has two basic parameters to adapt the segmentation to an specific problem, namely, h s , which controls a smoothing process, and h r related with the size of the kernel used to determine the modes and its basis of attraction. To test the method, we have selected the set parameters (h s , h r ) = {(7, 3), (7,15), (7,19), (7,23), (13, 7)(13, 19), (17,23)} given in [24] and [5]. The average times for this set of parameters, expressed in seconds, are 3.17, 4.15, 3.99, 4.07, 9.72, 9.69, 13.96 respectively.…”

Section: Results and Performance Evaluationmentioning

confidence: 99%

See 2 more Smart Citations

Image Segmentation in the Presence of Shadows and Highlights

Vazquez

Weijer

Baldrich

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The segmentation method proposed in this paper is based on the observation that a single physical reflectance can have many different image values. We call the set of all these values a dominant colour. These variations are caused by shadows, shading and highlights and due to varying object geometry. The main idea is that dominant colours trace connected ridges in the chromatic histogram. To capture them, we propose a new Ridge based Distribution Analysis (RAD) to find the set of ridges representative of the dominant colour. First, a multilocal creaseness technique followed by a ridge extraction algorithm is proposed. Afterwards, a flooding procedure is performed to find the dominant colours in the histogram. Qualitative results illustrate the ability of our method to obtain excellent results in the presence of shadow and highlight edges. Quantitative results obtained on the Berkeley data set show that our method outperforms state-of-the-art segmentation methods at low computational cost.

show abstract

Section: Results and Performance Evaluationmentioning

confidence: 99%

Section: Results and Performance Evaluationmentioning

confidence: 99%

Section: Results and Performance Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Image Segmentation in the Presence of Shadows and Highlights

Vazquez

Weijer

Baldrich

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…shows an example of mean shift initial segmentation. For detailed information about mean shift and EDISON system, please refer to [18,19,25,26]. In this paper we only focus on the region merging.…”

Section: Similarity Region Mergingmentioning

confidence: 99%

Unsupervised Method of Object Retrieval Using Similar Region Merging and Flood Fill

Saxena¹,

Jain²,

Singh³

2011

SpecialIssue

View full text Add to dashboard Cite

Abstract-In this work; we address a novel interactive framework for object retrieval using unsupervised similar region merging and flood fill method which models the spatial and appearance relations among image pixels. Efficient and effective image segmentation is usually very hard for natural and complex images. This paper presents a new technique for similar region merging and objects retrieval. The users only need to roughly indicate the after which steps desired objects boundary is obtained during merging of similar regions. A novel similarity based region merging mechanism is proposed to guide the merging process with the help of mean shift technique. A region R is merged with its adjacent regions Q if Q has highest similarity with R among all Q's adjacent regions. The proposed method automatically merges the regions that are initially segmented through mean shift technique, and then effectively extracts the object contour by merging all similar regions. Extensive experiments are performed on 22 object classes (524 images total) show promising results.

show abstract

“…The third segmentation was obtained using the synergistic segmentation of Christoudias et al [7] (figure 3(c)). The algorithm is an improvement of the mean-shift color segmentation [8] with the addition of an edge confidence parameter [19] in the feature space.…”

Section: Image Segmentationmentioning

confidence: 99%

Segmentation-based multi-class semantic object detection

Vieux

Benois-Pineau

Domenger

et al. 2010

Multimed Tools Appl

View full text Add to dashboard Cite

In this paper we study the problem of the detection of semantic objects from known categories in images. Unlike existing techniques which operate at the pixel or at a patch level for recognition, we propose to rely on the categorization of image segments. Recent work has highlighted that image segments provide a sound support for visual object class recognition. In this work, we use image segments as primitives to extract robust features and train detection models for a predefined set of categories. Several segmentation algorithms are benchmarked and their performances for segment recognition are compared. We then propose two methods for enhancing the segments classification, one based on the fusion of the classification results obtained with the different segmentations, the other one based on the optimization of the global labelling by correcting local ambiguities between neighbor segments. We use as a benchmark the Microsoft MSRC-21 image database and show that our method competes with the current state-of-the-art.

show abstract

Synergism in low level vision

Cited by 286 publications

References 14 publications

Image Segmentation in the Presence of Shadows and Highlights

Image Segmentation in the Presence of Shadows and Highlights

Unsupervised Method of Object Retrieval Using Similar Region Merging and Flood Fill

Segmentation-based multi-class semantic object detection

Contact Info

Product

Resources

About