Otsu Clustering Research Articles

Automatic Semantic Segmentation of Benthic Habitats Using Images from Towed Underwater Camera in a Complex Shallow Water Environment

Underwater image segmentation is useful for benthic habitat mapping and monitoring; however, manual annotation is time-consuming and tedious. We propose automated segmentation of benthic habitats using unsupervised semantic algorithms. Four such algorithms––Fast and Robust Fuzzy C-Means (FR), Superpixel-Based Fast Fuzzy C-Means (FF), Otsu clustering (OS), and K-means segmentation (KM)––were tested for accuracy for segmentation. Further, YCbCr and the Commission Internationale de l’Éclairage (CIE) LAB color spaces were evaluated to correct variations in image illumination and shadow effects. Benthic habitat field data from a geo-located high-resolution towed camera were used to evaluate proposed algorithms. The Shiraho study area, located off Ishigaki Island, Japan, was used, and six benthic habitats were classified. These categories were corals (Acropora and Porites), blue corals (Heliopora coerulea), brown algae, other algae, sediments, and seagrass (Thalassia hemprichii). Analysis showed that the K-means clustering algorithm yielded the highest overall accuracy. However, the differences between the KM and OS overall accuracies were statistically insignificant at the 5% level. Findings showed the importance of eliminating underwater illumination variations and outperformance of the red difference chrominance values (Cr) in the YCbCr color space for habitat segmentation. The proposed framework enhanced the automation of benthic habitat classification processes.

Automatic segmentation of nasopharyngeal carcinoma from CT images

The nasopharyngeal carcinoma (NPC) called also Cavum cancer becomes a public health problem for the Maghreb countries and Southeast Asia. The detection of this cancer could be carried out from computed tomography (CT) scans. In this context, we proposed two approaches based on image clustering to locate affected tissues by the Cavum cancer. These approaches are based respectively on E-M and Otsu segmentation. Compared to the physician manual contouring, our automatic detection proves that the detection of the cancer while using the Otsu clustering in terms of precision, recall and F-measure is more efficient than E-M. Then, we merged the results of these two methods by using the AND and the OR logical operators. The AND fusion yields to an increase of the precision while the OR fusion raises the recall. However, the detection of the NPC using Otsu remain the best solution in terms of F-Measure. Compared to previous studies that provide a surface analysis of the NPC, our approach provides a 3D estimation of this tumour ensuring a better analysis of the patient folder.

TUMOR CLASSIFICATION USING AUTOMATIC MULTI-THRESHOLDING

AbstractIn this paper we explore these math approaches for medical image applications. The application of the proposed method for detection tumor will be able to distinguish exactly tumor size and region. In this research, some major design and experimental results of tumor objects detection method for medical brain images is developed to utilize an automatic multi-thresholding method to handle this problem by combining the histogram analysis and the Otsu clustering. The histogram evaluations can decide the superior number of clusters firstly. The Otsu classification algorithm solves the given medical image by continuously separating the input gray-level image by multi-thresholding until reaching optimal smooth rate. The method solves exactly the problem of the uncertain contoured objects in medical image by using the Otsu clustering classification with automatic multi-thresholding operation.

Automatic segmentation of Nasopharyngeal carcinoma from CT images

The nasopharyngeal carcinoma (NPC) called also Cavum cancer becomes a public health problem for the Maghreb countries and Southeast Asia. The detection of this cancer could be carried out from computed tomography (CT) scans. In this context, we proposed two approaches based on image clustering to locate affected tissues by the Cavum cancer. These approaches are based respectively on E-M and Otsu segmentation. Compared to the physician manual contouring, our automatic detection proves that the detection of the cancer while using the Otsu clustering in terms of precision, recall and F-measure is more efficient than E-M. Then, we merged the results of these two methods by using the AND and the OR logical operators. The AND fusion yields to an increase of the precision while the OR fusion raises the recall. However, the detection of the NPC using Otsu remain the best solution in terms of F-Measure. Compared to previous studies that provide a surface analysis of the NPC, our approach provides a 3D estimation of this tumour ensuring a better analysis of the patient folder.

Methods toward in vivo measurement of zebrafish epithelial and deep cell proliferation

We present a strategy for automatic classification and density estimation of epithelial enveloping layer (EVL) and deep layer (DEL) cells, throughout zebrafish early embryonic stages. Automatic cells classification provides the bases to measure the variability of relevant parameters, such as cells density, in different classes of cells and is finalized to the estimation of effectiveness and selectivity of anticancer drug in vivo. We aim at approaching these measurements through epithelial/deep cells classification, epithelial area and thickness measurement, and density estimation from scattered points. Our procedure is based on Minimal Surfaces, Otsu clustering, Delaunay Triangulation, and Within-R cloud of points density estimation approaches. In this paper, we investigated whether the distance between nuclei and epithelial surface is sufficient to discriminate epithelial cells from deep cells. Comparisons of different density estimators, experimental results, and extensively accuracy measurements are included.