Shape Extraction Research Articles

DBSCAN-based point cloud extraction for Tomographic synthetic aperture radar (TomoSAR) three-dimensional (3D) building reconstruction

ABSTRACT Tomographic synthetic aperture radar (TomoSAR) has been widely used in three-dimensional (3D) reconstruction of urban buildings. However, due to the baseline distribution and the limitations of the algorithm itself, the building point cloud after tomographic imaging is flooded by substantial noise and/or false targets. Thus, TomoSAR point clouds must be extracted from these unwanted factors to reconstruct the building structure. Existing line-based extraction methods can only detect straight lines, which results in the loss of non-linear point clouds. Thus, inspired by density clustering, we propose a point cloud extraction method using the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. The DBSCAN can preserve the building structure more completely by enabling the extraction of various shapes of the buildings. Since the detection of point clouds is density-based, noise and false targets that exhibit low-density distribution can be accurately detected and rejected. The experimental results demonstrated the effectiveness of our method for TomoSAR point cloud extraction, as well as the structural protection of buildings, which achieves a higher extraction accuracy compared to linear detection.

Nipple Localization in Automated Whole Breast Ultrasound Coronal Scans Using Ensemble Learning.

Nipple is a vital landmark in the breast lesion diagnosis. Although there are advanced computer-aided detection (CADe) systems for nipple detection in breast mediolateral oblique (MLO) views of mammogram images, few academic works address the coronal views of breast ultrasound (BUS) images. This paper addresses a novel CADe system to locate the Nipple Shadow Area (NSA) in ultrasound images. Here the Hu Moments and Gray-level Co-occurrence Matrix (GLCM) were calculated through an iterative sliding window for the extraction of shape and texture features. These features are then concatenated and fed into an Artificial Neural Network (ANN) to obtain probable NSA's. Later, contour features, such as shape complexity through fractal dimension, edge distance from the periphery and contour area, were computed and passed into a Support Vector Machine (SVM) to identify the accurate NSA in each case. The coronal plane BUS dataset is built upon our own, which consists of 64 images from 13 patients. The test results show that the proposed CADe system achieves 91.99% accuracy, 97.55% specificity, 82.46% sensitivity and 88% F-score on our dataset.

Extraction of the structural mode shapes utilizing image processing method and data fusion

Dynamic properties of structures, such as natural frequency, mode shapes, and damping ratios, play a decisive role in the structural behavior against dynamic loads such as earthquakes. One of the utilizations of these properties is the determination of the earthquake forces imposed on structures based on the design spectra. Other applications of these features could be utilized to update the finite element model, detect damage in structures, and long-term health-monitoring of the structures. So, accuracy and reducing the cost of evaluating these properties would have a significant role in increasing the efficiency and the consequences of the useful life of the structures. Image processing is one of the methods for determination of dynamic properties of structures. The aim of this article is presenting a data fusion method to improve the accuracy of extraction of the mode shapes by image processing. For this purpose, a cantilever aluminum beam and a Three-story frame with additional mass have been considered, then their vibration was recorded utilizing the image processing method. To extract dynamic properties of structure, peak survey method was used. The proposed average method in signal level and Improved Distance Evaluation (IDE) in feature level were utilized to improve the accuracy of obtained mode shapes. The accuracy of the novel technique was investigated by comparing the seismic properties to the results from the accelerometer sensors and the finite element method. The results show that this method could enhance the accuracy of the image processing method for remote sensing.

COMPARISON BETWEEN AIRBORNE LIDAR DATA AND DIGITAL IMAGE MATCHING FOR SURFACE RECONSTRUCTION

Abstract. Nowadays some aerial surveying projects (both manned or not) might integrate imagery and ranging sensor technology, thus allowing both the reconstruction of the surface from airborne laser scanning (ALS) and from the Photogrammetric pipeline through digital image matching (DIM). DIM algorithms have been continuously improved, therefore comparisons to other techniques such as ALS must be conducted. Despite the scientific community efforts, there are few evaluations regarding the extraction of building roofs using point clouds derived from both ALS and imagery. In this sense, this study provides a brief comparison between ALS and DIM point clouds and the accuracy that can be achieved in surface reconstruction and building roofs extraction. The experiments indicated that in some situations such as the building roof shape extraction, the accuracy is similar between these techniques, however, this is not valid for vertical accuracy, where larger differences were observed.

FUZZY EDGE IMAGE MATCHING ALGORITHM FOR SQUID SPECIES IDENTIFICATION

Abstract: The image features play important role in matching system. The effectiveness of these Squid species features depends on the global features. The identification of Squid species requires information of their morphology. Body shape is very useful to characterize the one species to another species. In Shape extraction, edge detection is an important aspect. Edge is an important visual feature and it represents visual information with a limited number of pixels. While considering the morphology of Squid, it can have uncertainty due to climatic conditions. Hence, in this study feature extraction is done by fuzzy edge map. In this paper we proposed Fuzzy Image Edge Image Matching Algorithm (FEIMA) for Squid species identification. Similarity metric is used for matching of query and the candidate images in the database and it finally displays the class of species. The proposed algorithm performance is calculated by using Average of precision and recall.

GFNB: Gini index-based Fuzzy Naive Bayes and blast cell segmentation for leukemia detection using multi-cell blood smear images.

The blood cell counting and classification ensures the evaluation and diagnosis of a number of diseases. The analysis of white blood cells (WBCs) permits us to detect the acute lymphoblastic leukemia (ALL), a type of blood cancer that causes fatality when untreated. At present, the morphological analysis of blood cells is performed manually by skilled operators, which holds numerous drawbacks. The manual techniques for leukemia detection are time-consuming and show less accurate results. Hence, there is a need for an automatic method for detecting leukemia. In order to overcome the demerits associated with the manual methods of counting and classifying, an automatic method of blast cell counting and leukemia classification is progressed. This paper proposes a leukemia detection method, using the Gini index-based Fuzzy Naive Bayes (GFNB) classifier that is the integration of Gini index and Fuzzy Naive Bayes classifier. Initially, the input multi-cell blood smear image is subjected to pre-processing, and the blast cell is segmented using the adaptive thresholding. Then, the blast cells are counted using the proposed classifier so as to decide the presence of leukemia for the effective diagnosis. Experimental analysis using the ALL-IDB1 database confirms that the proposed method operates better than the existing methods in terms of accuracy, specificity, and sensitivity that are found to be 0.9591, 0.9599, and 1, respectively. The experimental results reveal that the proposed method is reliable and accurate. Also, the proposed system can help the physicians to improve and speed up their process.Graphical abstract Leukemia is caused by the excess production of the immature leucocytes in the bone marrow that expose the human body to lose the tendency to fight against the diseases. Leukemia classification is highly needed as in the later stage, failure of the diagnosis steps may lead to the death of the person. Moreover, some countries do not have any study against the diagnosis steps of leukemia and it highly exists among the low-income people. In order to analyze the type of leukemia and to provide an effective diagnosis strategy, the paper presents a fast and highly accurate classification method. The main aim of the paper is to propose a method to perform the leukemia classification through the segmentation and classification of the WBC cells using the multi-cell blood smear images. The major steps involved in the leukemia classification are pre-processing, segmentation, feature extraction, and classification. The input blood smear image is enhanced in the pre-processing step and the pre-processed image is subjected to segmentation using the LUV color transformation and Adaptive Thresholding strategy. The features are extracted from the individual segments and they are presented to the classifier for the classification. The features extracted are shape, texture, and count of the blast cells, for which the grid-based shape extraction, local gradient pattern (LGP)-based texture features, and pixel threshold-based counting of the blast cells are employed. The proposed classifier is developed using the Gini index and Fuzzy Naive Bayes classifier.

Extracting mode shapes from drive-by measurements to detect global and local damage in bridges

This article uses extracted bridge mode shapes to detect bridge damage, including both local damage such as that due to a bridge strike and global damage such as that due to foundation scour. Typically, only one accelerometer is required, which is mounted on an instrumented vehicle passing over the test bridge. This concept is referred to as ‘drive-by’ bridge health monitoring. For a simple vehicle/bridge interaction (VBI) model, where the vehicle is represented as a moving sprung mass, a closed-form solution for the bridge mode shape is derived from the acceleration response of the instrumented vehicle. The validity of the proposed algorithm is numerically verified using a VBI model. In addition, it is shown that the extracted mode shape can be used to assess the bridge condition. A damage index based on the extracted mode shape is proposed to detect bridge damage and to provide an indication of the damage level. In the numerical study, the factors that affect the accuracy of the proposed algorithm are investigated, including the vehicle speed and measurement noise. Two laboratory experiments are conducted to validate the mode shape extraction and bridge damage detection processes for cases of both local and global damage.

BOUNDARY REGULARIZED BUILDING FOOTPRINT EXTRACTION FROM SATELLITE IMAGES USING DEEP NEURAL NETWORKS

Abstract. In recent years, an ever-increasing number of remote satellites are orbiting the Earth which streams vast amount of visual data to support a wide range of civil, public and military applications. One of the key information obtained from satellite imagery is to produce and update spatial maps of built environment due to its wide coverage with high resolution data. However, reconstructing spatial maps from satellite imagery is not a trivial vision task as it requires reconstructing a scene or object with high-level representation such as primitives. For the last decade, significant advancement in object detection and representation using visual data has been achieved, but the primitive-based object representation still remains as a challenging vision task. Thus, a high-quality spatial map is mainly produced through complex labour-intensive processes. In this paper, we propose a novel deep neural network, which enables to jointly detect building instance and regularize noisy building boundary shapes from a single satellite imagery. The proposed deep learning method consists of a two-stage object detection network to produce region of interest (RoI) features and a building boundary extraction network using graph models to learn geometric information of the polygon shapes. Extensive experiments show that our model can accomplish multi-tasks of object localization, recognition, semantic labelling and geometric shape extraction simultaneously. In terms of building extraction accuracy, computation efficiency and boundary regularization performance, our model outperforms the state-of-the-art baseline models.

Automatic room information retrieval and classification from floor plan using linear regression model

The automatic creation of a repository of the building’s floor plan helps a lot to the architects to reuse them. The basic approach is to extract and recognize texts, symbols or graphics to retrieve the information of the floor plan from the images. This paper proposes a floor plan information retrieval algorithm. The proposed algorithm is based on shape extraction and room identification. $$\alpha $$ -shape is used for finding an accurate shape. From the detected shapes, actual areas of rooms are calculated. Later, a regression model-based binary room classification model is proposed to classify them into room-type, i.e., bedroom, drawing room, kitchen, and non-room-type, i.e., parking porch, bathroom, study room and prayer room. The proposed model is tested on the CVC-FP dataset with an average room detection accuracy of 85.71% and room recognition accuracy of 88%.

Using shape extraction to enhance classification of Landsat satellite images to visualize vegetation

Identifying vegetation from satellite images is very necessary for studies related to agricultural sector. Enhancing the quality and classification of satellite images is a challenge especially when study areas contain complex vegetation. The study aimed to enhance the image classification based on using extraction tool of the ArcGIS. In the present study, the classification of Landsat satellite images before and after extracting specific areas covering vegetation using the ArcGIS was compared. Two regions in Egypt, one with complex vegetation located at Siwa Oasis and the other one with simple vegetation located at Abu Simbel region, were used in this study. Then, several polygons were used to extract specific areas covering vegetation from the satellite images. Images were subsequently classified using Iso cluster unsupervised classification tool. The classification outcomes were compared between the original images before extraction with images after extraction. The results showed that extraction tool greatly enhanced the quality and classification of the image and it was possible to identify different types of vegetation.

Klasifikasi Obat Medis Berdasarkan Ekstraksi Ciri Menggunakan K-Means Clustering

The use of drugs and chemicals that should be used in the scope of the pharmaceutical field, but since 2014 the use of these drugs and chemicals has been misused in the community such as PCC (paracetamol), caffeine, carisoprodol, and carnophen. One of the causes of this abuse is due to the community's ignorance in the use and types of drugs, such as not knowing the meaning and meaning of the drug logo or special markings listed on the drug packaging. To make it easier to read and know the names of these types of logos or special marks, it is necessary to develop an application that is able to read logo patterns or special marks using digital image processing. In this study, an application was designed to classify types of medical drug patterns using k-means clustering based on the extraction of shape and texture features. The study began with taking a drug pattern image using a smartphone camera with details of 16 sample images and 24 test images. The image of this result, carried out cropping on the logo to get a detailed image. Furthermore, cropping images which are still in the form of RGB are converted into binary images and grayscale images for feature extraction processes. There are 6 feature parameters used, namely metric, eccentricity, contrast, correlation, energy, and homogeneity. The test results on the test image data, the percentage of success of 91.5% with an accuracy rate of 100% obtained in the test data types of OTC drugs, hard drugs, and narcotics, while OTC images are limited to 66%.

Probabilistic object detection and shape extraction in remote sensing data

Remote sensing mainly focuses on information extraction from data acquired by sensors on satellite and aerial platforms. Here, one such area of interest is ground object detection and shape extraction. Recently launched satellites and conventional aerial platforms (such as commercial UAV and professional drones) have sensors leading to more detailed and rich data source for this purpose. From these, data most of the times come in the form of optical images and LiDAR measurements. Resolution of this acquired data has increased significantly such that most ground objects (as buildings, trees, ships, cars, airplanes) can be detected and analyzed in detail. Therefore, computer vision methods have become extremely useful in remote sensing applications such as building detection and shape extraction for urban planning; tree crown measurement for crop yield forecasting; ship detection for monitoring unlawful fishery; car detection for traffic flow monitoring and intelligent transportation; and airplane detection for military and commercial operations. Researchers proposed several methods to automate the mentioned applications since manually handling them is extremely hard and prohibitively time consuming. Unfortunately, the proposed methods focus on one object type most of the times. Therefore, there is no general method to handle all the mentioned applications using computer vision tools. To overcome this problem, we propose a general framework for object detection and shape extraction in remote sensing data. Our method is based on probabilistic representation inspired by our previous work and perceptual organization principles. Due to space limitations, we only focus on buildings, trees, ships, airplanes, and cars as objects of interest in this study. We test the proposed method on several optical images acquired by different satellites and LiDAR data obtained from an aerial platform. For all objects of interest, we provide test results on both object detection and shape extraction steps. We analyze the proposed method based on these tests and discuss its strengths and weaknesses. We also comment on possible future extensions of the proposed method.

Morphological Detection and Extraction of Rail Surface Defects

Rail inspection by means of a visual system has been a subject of a number of publications in recent years. The main requirements with regard to such a system are that it has to be fast, nondestructive, and accurate. This article presents a system for rail defect detection and shape extraction utilizing morphological operations. The proposed system is evaluated using publicly available rail surface discrete defects (RSDDs) data sets of Type-I and Type-II. Several versions of the system are compared, and the best one is reduced to just 17 basic morphological operations implementing defect detection, defect extraction, and false-positive (FP) reduction scheme. Two evaluation methods are considered, which is a pixel-level method and a defect-level method. 70% precision and 79% recall were obtained at the defect level for Type-I RSDD. Similarly, 82% precision and 89% recall were obtained at the defect level for Type-II RSDD. Experiments showed that the FP reduction scheme is effective for type II RSDD when considering defect-level results, and it increases precision by 7%–9% while maintaining the recall almost intact. The execution time of the morphological operations of the detection algorithm together with FP reduction is in the order of 50 ms on a usual laptop.

Markov Random Field Based Color Texture Segmentation

Texture segmentation is one of the popular research domains and researchers across the globe are working on texture segmentation to enhance segmentation performance to address its requirements in many fields. Color texture segmentation has wide spectrum of applications in diverse fields such as segmentation of natural images, medical image analysis, remote sensing, shape extraction and inspection of products etc. This paper presents color texture segmentation algorithm which can satisfy requirements for such applications. Proposed algorithm is based on Markov Random Field (MRF) model eliminating the need of major contributor viz. Gabor filter used in past four decades for feature extraction and use only color as texture feature. Highly crude segmentation results are produced using only color as texture features. Crude segmentation results are enhanced by using Median filter with enlarged window size quantitatively determined by using parameters viz. structural similarity index (SSIM), mean square error (MSE) and peak signal to noise ratio (PSNR). Feature space dimensions are reduced by factor of 11 in proposed approach and this reduced computations by a factor of 11. The experimentation is carried out on 80 multi-class color texture benchmark images from Prague texture segmentation dataset and 4 benchmark images in Vistex dataset. Mean segmentation accuracy achieved for Prague texture dataset is 87.55% and it is higher by 9.82% over the best performing algorithm among 11 state-of-art algorithms suggested in most recent literature. Accuracy achieved for Vistex dataset is 98.21%. Average SSIM for Prague dataset is 0.91403 and Vistex dataset is 0.9405.

Classification of eye-fundus images with diabetic retinopathy using shape based features integrated into a convolutional neural network

The paper proposes a CNN architecture in which features extracted using shape from eye-fundus images are fed as an input. Firstly, we preprocess the images extensively. These preprocessed images are thereby used in the extraction of shape based features. The feature set obtained is then fed into Convolutional neural network (CNN) architecture for classification of retinopathy images into different groups. The proposed feature set shows an improvement in Average Retrieval Rate (ARR) up to 19.88% and Average retrieval precision (ARP) up to 12.01% over the vgg-f architecture of deep networks used for classification.

Model independent peak fitting and uncertainty assignment for high-precision time-of-flight mass spectrometry.

A procedure is introduced to extract the time-of-flight probability density distribution for ions of one type using an accumulated ion peak as a template for use in the fitting of time-of-flight peaks of co-accumulated ions. We demonstrate the use of logarithmic splines and discuss the statistical criteria for the selection of the spline smoothing parameter to obtain a smooth function for approximating such an ion peak. Furthermore, a "Bootstrap" method is proposed to determine the uncertainty of the fitted time-of-flight values, which is calculated using the extracted peak shape via a repeated statistical experiment. The developed procedure is validated against time-of-flight data obtained in a previously reported experiment. It is shown that atomic masses derived from our newly proposed methodology are in good agreement with those from the 2016 Atomic Mass Evaluation and the values determined using an analytically defined function (exponential-Gaussian hybrid), approximately matching the observed peak shape. In contrast to the previously used approach, our newly proposed methodology is a much more flexible, robust, and universal method for peak shape approximation, which can be applied to any peak shape.

A Principal Component Analysis (PCA) based Method for Shape Extraction

A Principal Component Analysis (PCA) based Method for Shape Extraction

Automatic Feature Region Searching Algorithm for Image Registration in Printing Defect Inspection Systems

Image registration is a key step in printing defect inspection systems based on machine vision, and its accuracy depends on the selected feature regions to a great extent. Aimed at the current problems of low efficiency and crucial errors of human vision and manual selection, this study proposes a new automatic feature region searching algorithm for printed image registration. First, all obvious shapes are extracted in the preliminary shape extraction process. Second, shape searching algorithms based on contour point distribution information and edge gradient direction, respectively, are proposed. The two algorithms are combined to put forward a relatively effective and discriminative feature region searching algorithm that can automatically detect shapes such as quasi-rectangular, oval, and so on, as feature regions. The entire image and the subregional experimental results show that the proposed method can be used to extract ideal shape regions, which can be used as characteristic shape regions for image registration in printing defect detection systems.

Ambient and free-vibration tests to improve the quantification and estimation of modal parameters in existing bridges

The correct estimation of modal parameters has an important role in ensuring the structural reliability of existing bridges. Operational modal analysis provides rather accurate extraction of natural frequency and mode shapes, but the corresponding damping estimates are subjected to higher uncertainties. This paper reports the results of ambient and free-vibration tests performed on five different typologies of road and railway bridges: steel trusses, steel box, multi-girder reinforced concrete, masonry and reinforced concrete arch bridges. In one case, data acquired by a continuous structural health monitoring system over a period of 1 year are available. In another case, the passage of trains is exploited to perform free decays. The processing of ambient vibrations is performed through several frequency and time-domain identification techniques, whereas the measured free decays are analyzed by the logarithmic decrement method applied to the autocorrelation functions of the signals. Outcomes are compared, evaluating the quality of modal damping estimates and the accuracy of results. The influence on modal parameters extraction of structural typologies, length of acquired time histories and ambient noise, loading and environmental effects are studied. At the end, it is demonstrated that the estimation of modal damping is more reliable for flexible structures when SHM and free-vibration data are available.

Deep Learning Object-Impulse Detection for Enhancing Leakage Detection of a Boiler Tube Using Acoustic Emission Signal

Advances in technology have enhanced the ability to detect leakages in boiler tube components in thermal power plants. As a specific issue, the interaction between the coal fuel stream and the boiler tube membrane generates random and high-amplitude impulses, which negatively affect the measured acoustic emission (AE) signal from leakages. It is essential to detect and practically handle these kinds of impulses. Based on the object detection concept, this paper proposes an impulse detection methodology that employs deep learning flexible boundary regression (DLFBR). First, the shape extraction (SE) preprocessing technique is implemented to yield the shape signal, which contains intrinsic information about the impulse from the raw AE signal. Then, DLFBR extracts and generates both the feature map and the confidence mask from the shape signal to regress a boundary box, which specifies the position of the impulse. For illustration purposes, the proposed algorithm is applied to an experimental leakage detection dataset recorded from a subcritical boiler unit with a tube membrane. Experimental results show that the proposed method is effective for detecting impulses of leakage in a boiler tube testbed, providing 99.8% average classification accuracy.