Otsu Thresholding Algorithm Research Articles

Detection of early bruises in apples using hyperspectral imaging and an improved MobileViT network.

Apples are susceptible to postharvest bruises, leading to a shortened shelf life and significant waste. Therefore, accurate detection of apple bruises is crucial to mitigate food waste. This study proposed an improved lightweight network based on MobileViT for detecting early-stage bruises in apples, utilizing hyperspectral imaging technology from 397.66 to 1003.81nm. After acquiring hyperspectral images, the Otsu threshold algorithm was employed for mask extraction, and principal component analysis was used for feature image extraction. Subsequently, the improved MobileViT network (iM-ViT) was implemented and compared with traditional algorithms, utilizing depthwise separable convolutions for parameter reduction and integrating local and global features to enhance bruise detection capability. The results demonstrated the superior performance of iM-ViT in accurately detecting apple bruises, showing significant improvements. The F1 score and test accuracy for detecting apple bruises using iM-ViT reached 0.99 and 99.07%, respectively. The fivefold cross-validation strategy was used to assess the stability and robustness of iM-ViT, and ablation experiments were performed to explore the effects of depthwise separable convolutions and local features on parameter reduction and classification accuracy improvement for early-stage bruise detection in apples. The results demonstrated that iM-ViT effectively reduced parameters and improved the ability to detect early bruises in apples. PRACTICAL APPLICATION: This study proposed an improved lightweight network to detect early bruises in apples, providing a reference for quick detection of bruises caused in the production process. Potential insights into the nondestructive detection of apple bruises using lightweight networks have been presented, which might be applied to mobile or online devices.

Comparative Analysis of Water Body Extraction Accuracy Based on Thresholding Method

Abstract. With the continuous development of remote sensing technology, various methods for land water body extraction based on satellite remote sensing have emerged. The thresholding method, as a commonly used image segmentation technique, possesses advantages such as high efficiency and wide applicability, making it widely employed in water body extraction research. In this thesis, utilizing SPOT4 imagery, we conducted experimental comparisons of water body extraction using the Iteration thresholding algorithm, Kittler-Illingworth (KI) thresholding algorithm, and Otsu thresholding algorithm. The experimental results demonstrate that all three thresholding methods exhibit good performance in water body extraction, with the Kittler-Illingworth (KI) thresholding algorithm achieving the highest relative accuracy.

Electrical Tree Image De-Noising using Threshold Wavelet Transform and Wiener Filter

Electrical treeing occurred in solid dielectric materials, especially in electrical application with high voltage. The occurrence of electrical tree happens when high electric fields applied, causing tiny channels or paths to form. The main issue during the data collection process is the changes of lighting, making it difficult to study the tree's propagation length, fractal dimension, and growth rate due to corrupted images. This research aims to analyse electrical tree structure images in XLPE material using a CCD camera and develop image de-noising techniques to suppress noise on the electrical tree image. The performance was then analysed using the Otsu thresholding algorithm for accurate segmentation. The methodology was divided into four phases: sample preparation, experimental setup, image pre-processing in MATLAB, and testing four de-noising filters: Wiener, median, NLM, and Gaussian. The Wiener filter with higher PSNR, SNR, and RMSE was selected and using superimposed method, both threshold wavelet transforms and wiener was combined to eliminate the noise. Finally, the proposed method of superimposed was tested with the Otsu thresholding method to evaluate accuracy, sensitivity, and specificity of the combination filter. Based on the analysis of PSNR, SNR, and RMSE, the performance of the threshold wavelet and Wiener filter (TWWF) de-noising technique improves the image quality of the electrical tree structure. Thus, for the Otsu thresholding segmentation algorithm analysis, it also had the highest values in terms of accuracy, sensitivity, and specificity.

MEG AND PET IMAGES-BASED BRAIN TUMOR DETECTION USING KAPUR’S OTSU SEGMENTATION AND SOOTY OPTIMIZED MOBILENET CLASSIFICATION

Deep learning techniques have revolutionized medical image analysis recently, particularly in brain tumor (BT) detection. A comprehensive overview of the advancements and challenges associated with employing deep learning methodologies for the accurate and timely detection of BT. This work proposes a novel brain hybrid hexagonal mobile network (BH2Mnet) to identify benign and malignant tumors using MEG and PET images. An adaptive bilateral filter (ABF) is used as a de-noising for input images to eliminate noise artifacts. Eliminating the skull and outer cortical regions, a process known as "skull stripping" is utilized to enhance the number of training images. The de-noised images are segmented to detect the BT using Kapur's Otsu threshold (KOT) algorithm. Based on these segmented tumors, hexagonal feature sets with and without segmentation masks are produced using hybrid hexagonal features (HHF). Finally, the Sooty optimization-based MobileNet classifier is employed to classify the BT into benign, malignant cases. It was determined that the proposed BH2Mnet approach was 99.21 % accurate in classifying data. According to the proposed BH2Mnet NS-CNN, the total accuracy is enhanced by 1.67 %, 2.69 %, and 4.11 % compared to hybrid DAE, BFC, Deep CNN, and Neutrosophy.

Comparison of Synthetic Aperture Radar Sentinel-1 and ALOS-2 observations for lake monitoring

This work investigates the efficacy of L-band and C-band Synthetic Aperture Radar (SAR) sensors onboard ALOS-2 and Sentinel-1 satellites, as compared to optical sensors onboard Sentinel-2 satellite, for mapping open water of the Tri An reservoir, one of the largest artificial reservoirs in South Vietnam, during the 2016-2023 period. The Google Earth Engine (GEE) was the primary computing platform to pre-process all satellite observations. The Otsu threshold algorithm was employed to generate water/non-water maps derived from the VH- and HH-polarized backscatter coefficient data acquired by Sentinel-1 and ALOS-2 satellites and from the Modified Normalized Difference Water Index (MNDWI) data acquired by Sentinel-2 satellite, respectively. The findings reveal the stability of Tri An reservoir’s surface water extent from 2017 to 2022, followed by a significant decline of nearly 70% during the dry season of 2023 to approximately 100 km2. This substantial decrease can be explained by the impact of a robust El Niño phase occurring in the region simultaneously. Overall, there is a high consistency between results derived from SAR and optical sensors, but the correlation between Sentinel-1 and Sentinel-2 (R = 0.9774) was higher than that between ALOS-2 and Sentinel-2 (R = 0.9145). During the drought period, both C-band and L-band SAR sensors overestimate the reservoir’s surface water extent due to the similarity in their backscatter coefficient between water and dry flat soil surfaces. This misclassification is more pronounced in ALOS-2 data than Sentinel-1 data, suggesting that the C-band sensor is more suitable than the L-band sensor for mapping the lake’s open water areas.

Detection of Brain Tumour Via Reversing Hexagonal Feature Pattern for Classifying Double-Modal Brain Images

The diagnosis of brain tumours (BT) is time-consuming and heavily dependent on the radiologists' abilities. Multiple algorithms have been developed for detecting and classifying BT that are both accurate and fast. Recent years have seen an increase in the popularity of deep learning, especially when it comes to developing automated systems that can diagnose and segment BT more accurately and with less time. In this paper, a novel Brain Hexagonal Pattern Network (BHPN) has been proposed to classify the MEG and PET images into normal, benign and malignant tumours. For pre-processing, a bilateral filter is employed to remove noise artifacts from the collected MEG and PET images. To remove the outer cortical and skull region, skull stripping is used, to be implemented to raise the volume of the training datasets. The pre-processed images are segmented using the Otsu threshold algorithm to segment the BT. These segmented tumours are taken as input to the Reversing Hexagonal algorithm to generate the hexagonal feature sets with and without a segmentation mask. In order to categorize tumours into normal, benign and malignant cases, a Spiking Dilated Convolutional Neural Network (SDCNN) classifier system is implemented. The classification accuracy of the Proposed BHPN approach is 99.54%. The Proposed BHPN approach improves the overall accuracy by 1.49%, 2.52%, and 3.93% better than hybrid deep autoencoder (DAE) and Bayesian Fuzzy Clustering (BFC), Deep CNN, and Neutrosophy and Convolutional Neural Network (NS-CNN) respectively.

A novel chaotic weighted EHO-based methodology for retinal vessel segmentation

ABSTRACT Retinal image segmentation process deals withproblems like spurious vascularisation and thin vessel detection. In thispaper, a three-step methodology has been proposed for retinal vesselsegmentation. In first step, RGB to YIQ conversion is performed. In second step, Y component enhancement is performed. A novel Chaotic weighted Elephant HerdingOptimization (CWEHO) has been proposed to optimize the clip limit and blocksize values of Contrast Limited Adaptive Histogram Equalization (CLAHE). CWEHO-based CLAHE along with morphological operations, non-local means filter, and median filter is applied to enhanceretinal images. In third step, thin and thick vessel segmentation is performed.Top hat transformation, otsu thresholding algorithm, and vessel point selectionare applied for thick vessel extraction. The first-order Gaussian derivative in conjunction with the matchfilter is used to extract thin vessels. DRIVE and HRF datasets are used toassess the effectiveness of proposed methodology. The average values ofsegmentation accuracy, specificity, sensitivity, and Mathew CorrelationCoefficient (MCC) are observed to be 0.9650, 0.9895, 0.7007, 0.7650,respectively, for observer1 and 0.9696, 0.9912, 0.7390, 0.7901 for observer2using DRIVE dataset. Similarly, 0.9592, 0.9839, 0.6850, and 0.7116,respectively, metrics for HRF dataset. Compared to state-of-the-art methods, the proposed segmentation methodology providesbetter results .

Early detection of citrus anthracnose caused by Colletotrichum gloeosporioides using hyperspectral imaging

Citrus fruit are susceptible to Colletotrichum gloeosporioides infestation during postharvest and shelf storage. Early and accurate detection of citrus anthracnose is conducive for carrying out targeted pesticide control and mitigating the potential spread of the disease. An early citrus anthracnose detection method using hyperspectral imaging and machine learning techniques is proposed. The hyperspectral data of sound citrus fruits were first collected and served as healthy samples, which were then inoculated with C. gloeosporioides and were further divided into asymptomatic and symptomatic samples. To characterize the global and local grayscale differences of the susceptible samples in different band images, the mean spectrum of the region of interest (ROI) of each band image was extracted as the global spectral features; moreover, the ROI in each band was segmented into disjointed local regions using the contrast limited adaptive histogram equalization and the Otsu thresholding algorithm, where the mean spectrum of the local regions were extracted as the local spectral features, respectively. The global and local spectral features were then concatenated into fused spectral features. Finally, the performance of the fused features was evaluated using support vector machine (SVM), k-NN and random forest (RF). The results showed that, compared with the conventional spectral feature-based methods, the proposed fused spectral features combined with SVM obtained the optimal results, where the average detection accuracy reached 91.97%. Furthermore, after applying feature selection using the successive projections algorithm (SPA), the resultant dimensionally-reduced fused spectral features also obtained acceptable results, with an average detection accuracy of 91.04%.

An integrated convolutional neural network for classifying small pulmonary solid nodules.

Achieving accurate classification of benign and malignant pulmonary nodules is essential for treating some diseases. However, traditional typing methods have difficulty obtaining satisfactory results on small pulmonary solid nodules, mainly caused by two aspects: (1) noise interference from other tissue information; (2) missing features of small nodules caused by downsampling in traditional convolutional neural networks. To solve these problems, this paper proposes a new typing method to improve the diagnosis rate of small pulmonary solid nodules in CT images. Specifically, first, we introduce the Otsu thresholding algorithm to preprocess the data and filter the interference information. Then, to acquire more small nodule features, we add parallel radiomics to the 3D convolutional neural network. Radiomics can extract a large number of quantitative features from medical images. Finally, the classifier generated more accurate results by the visual and radiomic features. In the experiments, we tested the proposed method on multiple data sets, and the proposed method outperformed other methods in the small pulmonary solid nodule classification task. In addition, various groups of ablation experiments demonstrated that the Otsu thresholding algorithm and radiomics are helpful for the judgment of small nodules and proved that the Otsu thresholding algorithm is more flexible than the manual thresholding algorithm.

Recognition of Lung Nodules in Computerized Tomography Lung Images using a Hybrid Method with Class Imbalance Reduction

Lung cancer is among the deadly diseases affecting millions globally every year. Physicians' and radiologists' manual detection of lung nodules has low efficiency due to the variety of shapes and nodule locations. The paper aims to recognize the lung nodules in computerized tomography (CT) lung images utilizing a hybrid method to reduce the problem space at every step. First, the suggested method uses the fast and robust fuzzy c-means clustering method (FRFCM) algorithm to segment CT images and extract two lungs, followed by a convolutional neural network (CNN) to identify the sick lung for use in the next step. Then, the adaptive thresholding method detects the suspected regions of interest (ROIs) among all available objects in the sick lung. Next, some statistical features are selected from every ROI, and then a restricted Boltzmann machine (RBM) is considered a feature extractor that extracts rich features among the selected features. After that, an artificial neural network (ANN) is employed to classify ROIs and determine whether the ROI includes nodules or non-nodules. Finally, cancerous ROIs are localized by the Otsu thresholding algorithm. Naturally, sick ROIs are more than healthy ones, leading to a class imbalance that substantially decreases ANN ability. To solve this problem, a reinforcement learning (RL)-based algorithm is used, in which the states are sampled. The agent receives a larger reward/penalty for correct/incorrect classification of the examples related to the minority class. The proposed model is compared with state-of-the-art methods on the lung image database consortium image collection (LIDC-IDRI) dataset and standard performance metrics. The results of the experiments demonstrate that the proposed model outperforms its rivals.

RDLNN-based Image Forgery Detection and Forged Region Detection Using MOT

Image forgery detection TEMPhas become an emerging research area due to the increasing number of forged images circulating on the internet and other social media, which leads to legal and social issues. Image forgery detection includes the classification of an image as forged or authentic and as well as localizing the forgery wifin the image. In this paper, we propose a Regression Deep Learning Neural Network (RDLNN) based image forgery detection followed by Modified Otsu Thresholding (MOT) algorithm to detect the forged region. The proposed model comprises five steps that are preprocessing, image decomposition, feature extraction, classification and block matching. In the preprocessing step, the RGB images are converted to YCbCr color format. Then, the images are decomposed using the new Polar Dyadic Wavelet Transform (PDyWT), followed by the extraction of important features. The classification phase called RDLNN effectively classifies the normal image and the forged image. For localization of the forgery, the forged image is divided into a number of blocks, and then Genetic Three Step Search (GTSS) algorithm is exploited to identify the dissimilar blocks. To get the exact forged region in the image, the dissimilar blocks are analyzed by the Modified Otsu Thresholding (MOT) algorithm. The proposed algorithm is compared wif widely used image forgery detection algorithms. The results show that the proposed method improves the forgery detection accuracy and precision by at least 6.04% and 3.77%, respectively, as compared to the already existent techniques such as ANFIS, KNN, ANN, and SVM. Moreover, the training time of the proposed network is lower by at least 64.3 % than the above existing techniques

WaterMaskAnalyzer (WMA)—A User-Friendly Tool to Analyze and Visualize Temporal Dynamics of Inland Water Body Extents

Freely available satellite imagery from the EU Copernicus program can record water surfaces precisely and at high temporal resolution. This paper provides the development status of the open-source demo software “WaterMaskAnalyzer” (WMA) for the determination of water body extents. The application allows simple to use on-demand monitoring of inland water dynamics by the Otsu-thresholding algorithm that automatically classifies water bodies. The tool can answer various hydrological issues related to disaster and water management, nature conservation, or water body monitoring. The first results from investigations of the Sentinel-1 time series in VH polarization show high accuracies with R2 = 0.824 compared to in situ measurements for the Quitzdorf reservoir in Saxony, Germany. Small or indented-shaped water bodies, as well as those with forested riparian zones, such as the Cranzahl (VH: R2 = 0.102 and VV: R2 = 0.251) and Klingenberg reservoirs (VH: R2 = 0.091 and VV: R2 = 0.146), only achieve a low R2 for VV and VH polarization but receive equally low RMSEs of 0.045 km2 (Cranzahl) and 0.077 km2 (Klingenberg). By separating out outliers and using correction factors, fast improvements in the accuracies can be expected. For future improvements, alternate classification methods and diverse new ground-truth data lead us to expect the next big step in development.

Detection and localization of citrus fruit based on improved You Only Look Once v5s and binocular vision in the orchard.

Intelligent detection and localization of mature citrus fruits is a critical challenge in developing an automatic harvesting robot. Variable illumination conditions and different occlusion states are some of the essential issues that must be addressed for the accurate detection and localization of citrus in the orchard environment. In this paper, a novel method for the detection and localization of mature citrus using improved You Only Look Once (YOLO) v5s with binocular vision is proposed. First, a new loss function (polarity binary cross-entropy with logit loss) for YOLO v5s is designed to calculate the loss value of class probability and objectness score, so that a large penalty for false and missing detection is applied during the training process. Second, to recover the missing depth information caused by randomly overlapping background participants, Cr-Cb chromatic mapping, the Otsu thresholding algorithm, and morphological processing are successively used to extract the complete shape of the citrus, and the kriging method is applied to obtain the best linear unbiased estimator for the missing depth value. Finally, the citrus spatial position and posture information are obtained according to the camera imaging model and the geometric features of the citrus. The experimental results show that the recall rates of citrus detection under non-uniform illumination conditions, weak illumination, and well illumination are 99.55%, 98.47%, and 98.48%, respectively, approximately 2–9% higher than those of the original YOLO v5s network. The average error of the distance between the citrus fruit and the camera is 3.98 mm, and the average errors of the citrus diameters in the 3D direction are less than 2.75 mm. The average detection time per frame is 78.96 ms. The results indicate that our method can detect and localize citrus fruits in the complex environment of orchards with high accuracy and speed. Our dataset and codes are available at https://github.com/AshesBen/citrus-detection-localization.

Volleyball Video Moving Target Tracking and Detection Algorithm Based on Multisensor Information Fusion

With the improvement of the level of sports competition technology, pure sports game broadcasting can no longer meet the requirements of various users. Users urgently need new technologies to meet the needs of quickly obtaining game information. The 5G Internet of Things is rapidly rising; volleyball is also gradually into people’s life. To solve the above problems, the research on the volleyball video moving target tracking and detection algorithm based on 5G Internet of Things communication and artificial intelligence becomes very important. This article is aimed at comparing the tracking and detection algorithms of volleyball video moving targets through 5G Internet of Things communication and artificial intelligence. In the process of detecting moving targets, several commonly used background detection methods are comprehensively compared and analyzed, and a method based on adaptive background difference is proposed. The algorithm uses Gaussian mixture model for background modeling, uses the OTSU threshold algorithm to determine the segmentation threshold T , and proposes a background update method that combines the expected full statistics and the L -close window to accelerate the background update speed, so that a more accurate sports prospect target. Experimental results show that the algorithm can detect and track moving targets in volleyball videos well in a fixed scene or when moving targets are not blocked, with a detection accuracy of more than 92%.

An Efficient Medical Image Processing Approach Based on a Cognitive Marine Predators Algorithm

Image processing aims to enhance the image's quality such that it is simple for both people and robots to understand. Medical image processing and Biomedical signal processing have many conceptual similarities. Medical image processing involves evaluation, enhancement, and presentation. The focus of medical imaging is on obtaining photographs for both therapeutic and diagnostic reasons. In the existing Marine Predator Algorithm, different disadvantages are experienced when various automated optimization algorithms are used to the problem of ECG categorization. The proposed method follows the flow outlined here: data collection, image preprocessing using histogram equalization, segmentation using the Otsu threshold algorithm, feature extraction using the contour method, feature selection using the Neighborhood Component Analysis (NCA) algorithm, and Cognitive Marine Predator Algorithm (CMPA) as the proposed method. By using the Cognitive Marine Predators Algorithm (CMPA), base layers are fused to use the greatest feasible parameters, producing enhanced high-quality output images. Finally, the image processing performance is analyzed. The proposed approaches overcome the drawbacks of existing algorithms and increase the quality of medical images efficiently.

A method for detecting plastic waste floating using Sentinel 2 high spatial resolution image: a case study in the coastal area of Vietnam

Ocean plastic waste pollution is now becoming a serious environmental problem, especially for a country with a long coastline and wide sea like Vietnam. The remote sensing method is considered suitable and effective in early detection and classification of ocean plastic waste due to the difference in spectral reflectance of plastic waste compared to the surrounding sea. This paper presents the results of identification and classification of plastic mesh in coastal areas of Vietnam by using Sentinel 2 MSI high spatial resolution optical images. First, water was extracted from Sentinel 2 image by thresholding method on a near-infrared band. Then, the plastic mesh was identified and classified based on Float Debris Index (FDI) index using Otsu thresholding algorithm. In the study, spectral indices such as NDVI, NDWI were also used to improve the accuracy in classifying plastic mesh. In the study, Google high spatial resolution satellite images were also used to evaluate the accuracy of plastic mesh classification. The obtained results show that, in 02 test areas, the proposed method allows detecting plastic mesh with an accuracy of over 90 %. The results obtained in the study can be used to provide input information for models of forecasting and assessing the impact of ocean plastic waste pollution on coastal environments.

Comparative Analysis of Machine Learning Algorithms in Automatic Identification and Extraction of Water Boundaries

Monitoring open water bodies accurately is important for assessing the role of ecosystem services in the context of human survival and climate change. There are many methods available for water body extraction based on remote sensing images, such as the normalized difference water index (NDWI), modified NDWI (MNDWI), and machine learning algorithms. Based on Landsat-8 remote sensing images, this study focuses on the effects of six machine learning algorithms and three threshold methods used to extract water bodies, evaluates the transfer performance of models applied to remote sensing images in different periods, and compares the differences among these models. The results are as follows. (1) Various algorithms require different numbers of samples to reach their optimal consequence. The logistic regression algorithm requires a minimum of 110 samples. As the number of samples increases, the order of the optimal model is support vector machine, neural network, random forest, decision tree, and XGBoost. (2) The accuracy evaluation performance of each machine learning on the test set cannot represent the local area performance. (3) When these models are directly applied to remote sensing images in different periods, the AUC indicators of each machine learning algorithm for three regions all show a significant decline, with a decrease range of 0.33–66.52%, and the differences among the different algorithm performances in the three areas are obvious. Generally, the decision tree algorithm has good transfer performance among the machine learning algorithms with area under curve (AUC) indexes of 0.790, 0.518, and 0.697 in the three areas, respectively, and the average value is 0.668. The Otsu threshold algorithm is the optimal among threshold methods, with AUC indexes of 0.970, 0.617, and 0.908 in the three regions respectively and an average AUC of 0.832.

Research on Sheepskin Contour Extraction Method Based on Computer Vision Measurement Technology

Manual trimming of sheepskin is intensive labor, and the working environment is full of rotten smells. The tannery is facing increasingly severe recruitment difficulties. This paper uses computer vision technology to study automatic recognition of sheepskin contours, which is the basis for the subsequent automatic trimming of sheepskin. After observing and analyzing the raw sheepskin images collected by an industrial array camera, a method of sheepskin contour extraction based on computer vision measurement technology is proposed in this paper. This method uses the fast Otsu threshold algorithm based on the pixel set to perform binary image segmentation. Combined with morphological processing for edge defect filling and topology analysis of boundary contour tracking algorithm to extract maximum contour information, it has a pixel-level three-dimensional de-noising preprocessing function and can accurately extract the sheepskin contour in the raw sheepskin image. The experimental results show that using the fast Otsu threshold algorithm proposed in this paper for binary segmentation to extract sheepskin contours, the detection rate is nearly 160% faster than the traditional Otsu algorithm, the edge protection is better, the error segmentation is reduced by nearly 3% and it has good anti-noise performance. It can meet the industrial production requirements of subsequent automatic cutting of sheepskin.

Monitoring spatial-temporal dynamics of small lakes based on SAR Sentinel-1 observations: a case study over Nui Coc Lake (Vietnam)

For the first time, this study estimates the variation of surface water extent of Nui Coc Lake located in Thai Nguyen province in North Vietnam at high spatial (20 m) and temporal resolution (bi-weekly). The classification methodology was developed based on the use of the Otsu threshold algorithm on the histogram of the backscatter coefficient of the SAR Sentinel-1 signal. Totally, more than 150 SAR Sentinel-1 images have been processed for the 2016-2020 period. Except for extreme drought and flood conditions, the average minimum and maximum of the lake’s surface water extent are 17 km2 (in May) and 24 km2 (in September/October), respectively, and Nui Coc Lake’s surface water was stable during the last five years. Classification results are in good agreement with the corresponding surface water extent maps derived from free-cloud Sentinel-2 images, with the occurrence map derived from the Landsat-derived Global Surface Water (GSW) product, and with in situ precipitation data. Compared to Sentinel-2, the lake’s surface water extent detected from Sentinel-1 is 4-4.5% less. The water occurrence is similar between our results and that derived from the GSW product, but Sentinel-1 data provide more details as its spatial resolution is higher than Landsat. This study clearly shows the great potential of SAR Sentinel-1 data for monitoring small lake’s water surface at low costs, especially over tropical regions.

Extraction of Cerebral Hemorrhage on CT Images Using Level Set Algorithm and Otsu Threshold

Extraction of cerebral hemorrhage on CT images has always been the focus of several research hotspots and is still challenging as it does not show clear boundary. In this paper, a novel segmentation framework is presented for extracting the cerebral hemorrhage in brain CT images with weak boundary. Firstly, we utilize the Otsu threshold algorithm to get the coarse outline approximate to the target boundary as the initial curve of level set algorithm. Then, the active contour model is employed using both edge information and global Gaussian distribution fitting energy of images to modify energy function of level set. The proposed approach is applied on real images which from Quzhou People’s Hospital. Compared to manual delineation, the proposed technique shows a higher JS value than the existing methods and requires less interaction which is listed in the literature.