Spatial-based skin detection using discriminative skin-presence features

EURASIP J. Adv. Signal Process.

Nalepa

et al. 2014

Self Cite

In this paper, we introduce a new self-adaptive algorithm for segmenting human skin regions in color images. Skin detection and segmentation is an active research topic, and many solutions have been proposed so far, especially concerning skin tone modeling in various color spaces. Such models are used for pixel-based classification, but its accuracy is limited due to high variance and low specificity of human skin color. In many works, skin model adaptation and spatial analysis were reported to improve the final segmentation outcome; however, little attention has been paid so far to the possibilities of combining these two improvement directions. Our contribution lies in learning a local skin color model on the fly, which is subsequently applied to the image to determine the seeds for the spatial analysis. Furthermore, we also take advantage of textural features for computing local propagation costs that are used in the distance transform. The results of an extensive experimental study confirmed that the new method is highly competitive, especially for extracting the hand regions in color images.

Section: Hybrid Methodsmentioning

confidence: 99%

“…Subsequently, the local model is applied to locate the seeds for spatial analysis, which determines the final boundaries of the skin regions. We perform the spatial analysis using the discriminative skin-presence features (DSPF), introduced in our earlier work [9], that rely on textural properties of skin probability maps.…”

Section: Contributionmentioning

confidence: 99%

Self-adaptive algorithm for segmenting skin regions

EURASIP J. Adv. Signal Process.

Nalepa

et al. 2014

Self Cite

“…There have been a number of methods for skin detection and segmentation proposed [19][20][21] that are based on skin color modeling [18], supported with the analysis of the texture [22] as well as spatial distribution of skin pixels [44]. Also, adapting the skin model [21,62] to a presented scene increases the precision of segmenting skin regions.…”

Section: Gesture Recognition Processmentioning

confidence: 99%

“…Every image in the set presents a singlehand gesture, and the ground-truth data encompass the SPM and locations of 25 landmarks. Our data sets (HGR1 and HGR2) contain images acquired with controlled and uncontrolled background, which makes skin segmentation challenging [21,22]. In the database, we provide ground-truth SPMs-this makes it possible to validate skin segmentation algorithms, and also enables evaluating hand shape recognition [36] as well as landmark detection and localization relying on the ground-truth data, so as to prevent error propagation.…”

Section: Data Setsmentioning

confidence: 99%

Hand landmarks detection and localization in color images

Grzejszczak

Gałuszka

2015

Multimed Tools Appl

Self Cite

This paper introduces a new method for detecting and localizing hand landmarks in 2D color images. Location of the hand landmarks is an important source of information for recognizing hand gestures, effectively exploited in a number of recent methods which operate from the depth maps. However, this problem has not yet been satisfactorily solved for 2D color images. Here, we propose to analyze the skin-presence masks, as well as the directional image of a hand using the distance transform and template matching. This makes it possible to detect the landmarks located both at the contour and inside the hand masks. Moreover, we performed an extensive experimental study to compare the proposed method with a number of state-of-the-art algorithms. The obtained quantitative and qualitative results clearly indicate that our approach outperforms other methods, which may help improve the existing gesture recognition systems.

“…This step is undertaken if the shape features are to be extracted from the skin map of I i . There exist a number of robust skin detection and segmentation techniques [15][16][17][18][19]. A thorough survey on current state-of-the-art skin detection approaches has been published recently [20].…”

Section: Parallel Hand Shape Classification Algorithmmentioning

confidence: 99%

Fast and Accurate Hand Shape Classification

Nalepa

Communications in Computer and Information Science

2014

Self Cite

Abstract. The problem of hand shape classification is challenging since a hand is characterized by a large number of degrees of freedom. Numerous shape descriptors have been proposed and applied over the years to estimate and classify hand poses in reasonable time. In this paper we discuss our parallel framework for real-time hand shape classification applicable in real-time applications. We show how the number of gallery images influences the classification accuracy and execution time of the parallel algorithm. We present the speedup and efficiency analyses that prove the efficacy of the parallel implementation. Noteworthy, different methods can be used at each step of our parallel framework. Here, we combine the shape contexts with the appearance-based techniques to enhance the robustness of the algorithm and to increase the classification score. An extensive experimental study proves the superiority of the proposed approach over existing state-of-the-art methods.