A study of lesion skin segmentation, features selection and classification approaches

Filali, Youssef; Ennouni, Assia; Sabri, My Abdelouahed; Aarab, Abdellah

doi:10.1109/isacv.2018.8354069

Cited by 26 publications

(11 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“… Mathematical formulation of the main descriptors derived from GLCMs Although GLCMs provide a rich description of spatial dependence, it is impractical to manipulate them in their raw form. A set of 14 statistical descriptors or attributes are well known in the literature to summarize the textual information contained in the GLCMs, but the five descriptors appearing most often are entropy, energy, contrast, homogeneity, and correlation [38] [39].…”

Section: Techniques Of Texture Features  Gray Level Co-occurrencementioning

confidence: 99%

Withdrawn

2020

IJEAT

View full text Add to dashboard Cite

Today, skin cancer is a serious health problem and the quest for a correct clinical diagnosis has been an ongoing concern for dermatologists and for many researchers an important field study. Field Programmable Gate Array (FPGA) technology has become a key part of the current movement towards innovation. It provides a viable target for applying image processing algorithms. It presents a viable target for the implementation of algorithms suited to image processing even under real-time requirements, because of their flexibility and great potential. FPGAs have shown very high performance in spite of their low operational frequency and high parallelism in applications. In this paper, an embedded Content-Based Image Retrieval (CBIR) implementation based on FPGA is proposed to manage the image data efficiently and fastly. We describe the implementation of CBIR in image processing applications using FPGA. The main aim behind this implementation is to reach minimum timing by the maximum utilization of resources. We are conducting a study on the advantages of this embedded approach and discussed its effectiveness for a set of benchmark images publically available dataset PH2. The FPGA-based implementation shows the feasibility of the proposed methodology. The functional implementation of all processes using the Verilog code of FPGA has been compiled on the ALTERA QUARTUS-II software tool.

show abstract

Section: Techniques Of Texture Features  Gray Level Co-occurrencementioning

confidence: 99%

Withdrawn

2020

IJEAT

View full text Add to dashboard Cite

show abstract

“…Certainly each of them has its own weaknesses and their use depends on the types of data to classify, but it is also possible to find a good compromise and to use simple methods that lead to high performance with a fast computing time. In our work, the choice of a classification algorithm to use is based on our previous work (Filali et al, 2018) where the SVM with quadratic kernel gives the best result compared with other classifiers and kernels. The SVM is a supervised learning.…”

Section: Features Engineeringmentioning

confidence: 99%

“…CAD approaches for the diagnosis of skin lesions commonly use the previously described steps from pre-processing to lesion classification through segmentation and feature extraction. In this paper, we propose to use an image decomposition based on an algorithm derived from PDE, which has been proven both for image segmentation and for content retrieval of textured images (Filali et al, 2018;2017b). This algorithm yields two components.…”

Section: Introductionmentioning

confidence: 99%

New and Efficient Features for Skin Lesion Classification based on Skeletonization

Filali¹,

Khoukhi²,

Sabri³

et al. 2019

Journal of Computer Science

Self Cite

View full text Add to dashboard Cite

This paper presents a new approach to detect and classify skin lesions for melanoma diagnosis with high accuracy. Skin lesion detection is based on an image decomposition into two components using the Partial Differential Equation (PDE). The first component that sufficiently preserves the contour is thus exploited to have an adequate segmentation of image lesion while the second component provides a good characterization of the texture. Moreover, to improve the classification accuracy, new and powerful features extracted by skeletonization of the lesion are presented. These features are compared and combined with well-known features from the literature. Features engineering was applied to select the most relevant features to be retained for the classification phase. The proposed approach was implemented and tested on a large database and gave a good classification accuracy compared to recent approaches from the literature.

show abstract

“…After pre-processing of skin lesion images, the next process which plays a significant role in non-invasive diagnosis of skin lesion as well as classification of skin lesions as cancer or non-cancerous is image segmentation. If the segmentation of skin lesion is not appropriate, then the performance of classifier will have adverse effects on classification accuracy [8]. Various segmentation methods like otsu thresholding, watershed, fcm, level set, and active contour are used to segment (separate) the foreground (skin lesion) from the acquired image without the need of biopsy [6,8,9,10,11].…”

Section: Introductionmentioning

confidence: 99%

“…If the segmentation of skin lesion is not appropriate, then the performance of classifier will have adverse effects on classification accuracy [8]. Various segmentation methods like otsu thresholding, watershed, fcm, level set, and active contour are used to segment (separate) the foreground (skin lesion) from the acquired image without the need of biopsy [6,8,9,10,11]. The efficiency of segmentation method is evaluated through certain performance comparison parameters like Jaccard index, Dice index, Hammoude index, TDR, and FDR [1,9,10].…”

Section: Introductionmentioning

confidence: 99%

Benign and Malignant Dermatoscopy Image Classification

Bobby

2020

2020 Third International Conference on Advances in Electronics, Computers and Communications (ICAECC)

View full text Add to dashboard Cite

Skin cancer is one of the life threatening diseases and there is mostly no chance of remission from skin cancer if diagnosed in the last stage. The three major types of skin cancers are basal cell carcinoma, squamous cell carcinoma, and melanoma. Among these three, melanoma skin cancer is dangerous and acute in nature. Dermatologists use various techniques for diagnosing the malignant, in which the popular and reliable clinical method is dermatoscopy. The manual inference of the disease condition from the dermatoscopy images requires intensive knowledge and experience in the related field. Also, there will be an unavoidable degree of variability in image analysis occurs as long as the diagnostics procedure relies on human visual perception. Thus recently, image processing and machine learning algorithms have been applied for the accurate diagnoses of skin cancers from the dermatoscopic images. Thus, the goal of the proposed work is to automatically segment and classify the dermatoscopy skin lesion image with the help of image processing and machine learning algorithms. The proposed approach classifies the skin lesion image as benign or malignant melanoma with 90% accuracy, 91% sensitivity, 86% specificity, and 93% precision.

show abstract

A study of lesion skin segmentation, features selection and classification approaches

Cited by 26 publications

References 7 publications

Withdrawn

Withdrawn

New and Efficient Features for Skin Lesion Classification based on Skeletonization

Benign and Malignant Dermatoscopy Image Classification

Contact Info

Product

Resources

About