Seed Point Selection Research Articles

An Improved Adaptive Grid-Based Progressive Triangulated Irregular Network Densification Algorithm for Filtering Airborne LiDAR Data

Ground filtering is crucial for airborne Light Detection and Ranging (LiDAR) data post-processing. The progressive triangulated irregular network densification (PTD) algorithm and its variants outperform others in accuracy, stability, and robustness, using grid-based seed point selection, TIN construction, and iterative rules for ground point identification. However, these methods still face limitations in removing low points and accurately preserving terrain details, primarily due to their sensitivity to grid size. To overcome this issue, a novel PTD filtering algorithm based on an adaptive grid (AGPTD) was proposed. The main contributions of the proposed method include an outlier removal method using a radius outlier removal algorithm and Kd-tree, a method for establishing an adaptive two-level grid based on point cloud density and terrain slope, and an adaptive selection method for angle and distance thresholds in the iterative densification processing. The performance of the AGPTD algorithm was assessed based on widely used benchmark datasets. Results show that the AGPTD algorithm outperforms the classical PTD algorithm in retaining ground feature points, especially in reducing Type I error and average total error significantly. In comparison with other advanced algorithms developed in recent years, the novel algorithm showed the lowest average Type I error, the minimal average total error, and the greatest average Kappa coefficient, which were 1.11%, 2.28%, and 90.86%, respectively. Additionally, the average accuracy, precision, and recall of AGPTD were 97.69%, 97.52%, and 98.98%, respectively.

Computer-aided diagnosis of cystic lung diseases using CT scans and deep learning.

Auxiliary diagnosis of different types of cystic lung diseases (CLDs) is important in the clinic and is instrumental in facilitating early and specific treatments. Current clinical methods heavily depend on accumulated experience, restricting their applicability in regions with less developed medical resources. Thus, how to realize the computer-aided diagnosis of CLDs is of great clinical value. This work proposed a deep learning-based method for automatically segmenting the lung parenchyma in computed tomography (CT) slice images and accurately diagnosing the CLDs using CT scans. A two-stage deep learning method was proposed for the automatic classification of normal cases and five different CLDs using CT scans. Lung parenchyma segmentation is the foundation of CT image analysis and auxiliary diagnosis. To meet the requirements of different sizes of the lung parenchyma, an adaptive region-growing and improved U-Net model was employed for mask acquisition and automatic segmentation. The former was achieved by a self-designed adaptive seed point selection method based on similarity measurement, and the latter introduced multiscale input and multichannel output into the original U-Net model and effectively achieved the lightweight design by adjusting the structure and parameters. After that, the middle 30 consecutive CT slice images of each sample were segmented to obtain lung parenchyma, which was employed for training and testing the proposed multichannel parallel input recursive MLP-Mixer network (MPIRMNet) model, achieving the computer-aided diagnosis of CLDs. A total of 4718 and 16290 CT slice images collected from 543 patients were employed to validate the proposed segmentation and classification methods, respectively. Experimental results showed that the improved U-Net model can accurately segment the lung parenchyma in CT slice images, with the Dice, precision, volumetric overlap error, and relative volume difference of 0.96±0.01, 0.93±0.04, 0.05±0.02, and 0.05±0.03, respectively. Meanwhile, the proposed MPIRMNet model achieved appreciable classification effect for normal cases and different CLDs, with the accuracy, sensitivity, specificity, and F1 score of 0.8823±0.0324, 0.8897±0.0325, 0.9746±0.0078, and 0.8831±0.0334, respectively. Compared with classical machine learning and convolutional neural networks-based methods for this task, the proposed classification method had a preferable performance, with a significant improvement of accuracy of 10.74%. The work introduced a two-stage deep learning method, which can achieve the segmentation of lung parenchyma and the classification of CLDs. Compared to previous diagnostic tasks targeting single CLD, this work can achieve various CLDs' diagnosis in the early stage, thereby achieving targeted treatment and increasing the potential and value of clinical applications.

Brain Blood Vessel Segmentation based on Region Growing and U-net Neural Network

Aiming at the problems of insufficient feature extraction in existing deep learning-based cerebrovascular segmentation methods and uneven segmentation results when the original imaging quality is poor, a cerebrovascular segmentation method that integrates region growing and an improved U-net neural network is proposed. Based on the deep learning segmentation results, automatic seed point selection for region growing is performed and fused to improve the segmentation accuracy at the pixel level. In this work, the integrated segmentation method improves the Dice coefficient from 0.805 to 0.878 and the Average Hausdorff Distance (AHD) from 3.949 to 0.903 compared to the deep learning segmentation method.

Computer-Aided Design on Image Detection: A Chronology Review

In order to conduct research and diagnose diseases, image processing is essential. Medical professionals frequently segment images for pre- and post-surgery decisions, which are necessary for treatment planning. In the medical research field, segmentation is the core subject of several studies. Computer-aided detection (CAD) is utilised to achieve the highest level of classification accuracy and may be used to identify tissues growing abnormally. For the purpose of finding abnormalities, magnetic resonance imaging (MRI) is an effective approach, but it takes time and requires a fair amount of human resources. This approach, however, was problematic for slicing data related to the interior surfaces of cavity structures, for instance, the human skull. As a result, a ray casting algorithm was used to create a software programme. The most significant problem with segmentation techniques for x-ray images is seed point selection. An object's surface structure is described by a three-dimensional (3D) surface structure graph (SSG) that was created during segmentation. Ultrasound image detection is critical today. The model can be further modified using this CAD software so that it can be reproduced on a rapid prototyping device in the STL file format. The suggested deep learning method is exceptionally effective in accurately detecting faults in each layer, according to experimental data.

An optimized seed point selection method assists digital image correlation for measuring deformation from OCT images of fundus diseases

Fundus diseases are a significant cause of vision loss in older individuals worldwide. This study aimed to analyze deformations in fundus optical coherence tomography (OCT) images by utilizing and improving digital image correlation (DIC), which is expected to enhance the comprehensive assessment of these diseases. Seed points were identified at the intersections of blood vessel-induced shadows and the interfaces between retinal layers, providing accurate initial estimates for DIC. Despite the presence of severe decorrelation effects due to large deformation and low speckle quality, full-field deformation could be successfully measured. Verification tests were conducted using OCT images of a patient with retinal vein occlusion (RVO), and the method was subsequently applied to diabetic macular edema (DME) and polypoid choroidal vascular disease (PCV). DIC can obtain accurate deformation fields for fundus OCT images based on the proposed seed point selection method, offering critical imaging indicator for diagnosing and treating fundus diseases.

A New Strategy for Individual Tree Detection and Segmentation from Leaf-on and Leaf-off UAV-LiDAR Point Clouds Based on Automatic Detection of Seed Points

Accurate and efficient estimation of forest volume or biomass is critical for carbon cycles, forest management, and the timber industry. Individual tree detection and segmentation (ITDS) is the first and key step to ensure the accurate extraction of detailed forest structure parameters from LiDAR (light detection and ranging). However, ITDS is still a challenge to achieve using UAV-LiDAR (LiDAR from Unmanned Aerial Vehicles) in broadleaved forests due to the irregular and overlapped canopies. We developed an efficient and accurate ITDS framework for broadleaved forests based on UAV-LiDAR point clouds. It involves ITD (individual tree detection) from point clouds taken during the leaf-off season, initial ITS (individual tree segmentation) based on the seed points from ITD, and improvement of initial ITS through a refining process. The results indicate that this new proposed strategy efficiently provides accurate results for ITDS. We show the following: (1) point-cloud-based ITD methods, especially the Mean Shift, perform better for seed point selection than CHM-based (Canopy Height Model) ITD methods on the point clouds from leaf-off seasons; (2) seed points significantly improved the accuracy and efficiency of ITS algorithms; (3) the refining process using DBSCAN (density-based spatial clustering of applications with noise) and kNN (k-Nearest Neighbor classifier) classification significantly reduced edge errors in ITS results. Our study developed a novel ITDS strategy for UAV-LiDAR point clouds that demonstrates proficiency in dense deciduous broadleaved forests, and this proposed ITDS framework could be applied to single-phase point clouds instead of the multi-temporal LiDAR data in the future if the point clouds have detailed tree trunk points.

Fourfold Bounce Scattering-Based Reconstruction of Building Backs Using Airborne Array TomoSAR Point Clouds

Building reconstruction using high-resolution tomographic synthetic aperture radar (TomoSAR) point clouds has been very attractive in numerous applications, such as urban planning and dynamic city modeling. However, for side-looking TomoSAR, it is a challenge to reconstruct the obscured backs of buildings using traditional single-bounce scattering-based methods. It comes to our attention that the higher-order scattering points in airborne array TomoSAR point clouds may provide rich information on the backs of buildings. In this paper, the fourfold bounce (FB) scattering model of combined buildings in airborne array TomoSAR is derived, which not only explains the cause of FB scattering but also gives the distribution pattern of FB scattering points. Furthermore, a novel FB scattering-based method for the reconstruction of building backs is proposed. First, a two-step geometric constraint is used to detect the candidate FB scattering points. Subsequently, the FB scattering points are further detected by seed point selection and density estimation in the radar coordinate system. Finally, the backs of buildings can be reconstructed using the footprint inverted from the FB scattering points and the height information of the illuminated facades. To verify the FB scattering model and the effectiveness of the proposed method, the results from the simulated point clouds and the real airborne array TomoSAR point clouds are presented. Compared with the traditional roof point-based methods, the outstanding advantage of the proposed method is that it allows for the high-precision reconstruction of building backs, even in the case of poor roof points. Moreover, this paper may provide a novel perspective for the three-dimensional (3D) reconstruction of dense urban areas.

Guided neighborhood affine subspace embedding for feature matching

Feature matching, which refers to determining reliable correspondences between two sets of feature points, is a fundamental component of numerous visual tasks. This paper proposes a novel method, termed as guided neighborhood affine subspace embedding (NASE), to eliminate false matches from the given tentative feature matches. Its essential philosophy is to preserve the underlying intrinsic manifold of potential true matches. Specifically, we aim to approximate the manifold of an inlier with an affine subspace fitted on its neighbors by imposing a motion-consistency constraint. Considering that the “corresponding manifold” of inliers may be biased by gross outliers, we introduce a density-based seed point selection strategy for neighborhood refinement. Based on the above two strategies, we further formulate the general feature matching problem into a mathematical optimization model and deduce a closed-form solution with linearithmic time complexity (i.e., O(NlogN)) for mismatch removal. Additionally, we devise a multi-scale strategy for neighborhood construction, making our method more robust to various degradations. Extensive experiments on general feature matching, fundamental matrix estimation, and loop closure detection demonstrate the clear superiority of NASE over the state-of-the-arts.

An effective approach for CT lung segmentation using region growing

X-ray is an important means of detecting lung diseases. With the increasing incidence of lung diseases, computer-aided diagnosis technology is of great significance in clinical treatment. It has become a hot research direction to use computer-aided diagnosis to recognize chest radiography images, which can alleviate the uneven status of regional medical level. For clinical diagnosis, medical image segmentation can enable users to timely obtain the target region they are interested in and analyze it, which is significant to be used as an important basis for auxiliary research and judgment. In this case, a region growing algorithm based on threshold presegmentation is selected for lung segmentation, which integrates image enhancement, threshold segmentation, seed point selection and morphological post-processing, etc., to improve the segmentation effect, which also has certain reference value for other medical image processing.

A novel k-mean clustering based graph cut for brain MR image segmentation

Image Segmentation is an influential method to detect and distinguish the diseased and normal sections of an image. Image segmentation extracts the regions of interest for precise diagnosis of tumours and scheduling appropriate line of treatment. The distinct tumours in brain have varied outlies, position and intensity values. Consequently, it is difficult to develop a common technique for brain MRI segmentation. Moreover, the extraction of anomalies from the brain MRI turns out to be a challenging task. In this research, we have developed a self-regulating method for selection of seed points for partitioning the graph of MR images with tumor to attain graph cut segmentation. The proposes method overcomes the key problem of primary seed point selection by utilizing the balanced brain layout and combines k - mean clustering into graph cut for segmenting an image. The outcomes attained by the proposed technique facilitates improved segmentation of the diseased region.

Detection of Brain Tumor in MR images using hybrid Fuzzy C-mean clustering with graph cut segmentation technique

In the medical image processing field segmentation task is very important process in order to diagnose any diseases in medical images taken from different machine. Medical images from MRI, CT scan, X-rays, Ultrasound and PET have different features, so segmentation process is very challenging task. In this paper, the graph theoretical approaches are proposed because it has flexibility representing any complex structure. Before segmentation process, MR images are pre-processed using Region of Interest, Inverse method and Boundary detection method. In this method the segmentation of MRI brain images is performed using Fuzzy C-mean clustering with graph cut techniques. FCM algorithm is proved to be efficient in term of computational rate by improving cluster center and modified way to selection of seed points. Seed point selection is by using FCM clustering derives new technique called Fuzzy C-mean seed selection (FCMSS). Image segmentation is performed using obtained path in the graph applied on set of cluster region of interest on MR images to detect brain tumors.

Variants of bat algorithm for solving partitional clustering problems

Clustering is an exploratory data analysis technique that organize the data objects into clusters with optimal distance efficacy. In this work, a bat algorithm is considered to obtain optimal set of clusters. The bat algorithm is based on the echolocation feature of micro bats. Moreover, some improvements are proposed to overcome the shortcoming associated with bat algorithm like local optima, slow convergence, initial seed points and trade-off between local and global search mechanisms etc. An enhanced cooperative co-evolution method is proposed for addressing the initial seed points selection issue. The local optima issue is handled through neighbourhood search-based mechanism. The trade-off issue among local and global searches of bat algorithm is addressed through a modified elitist strategy. On the basis of aforementioned improvements, three variants (BA-C, BA-CN and BA-CNE) of bat algorithm is developed and efficacy of these variants is tested over twelve benchmark clustering datasets suing intra-cluster distance, accuracy and rand index parameters. Simulation results showed that BA-CNE variant achieves more effective clustering results as compared to BA-C, BA-CN and BA. The simulation results of BA-CNE are also compared with several existing clustering algorithms and two statistical tests are also applied to investigate the statistical difference among BA-CNE and other clustering algorithms. The simulation and statistical results confirmed that BA-CNE is an effective and robust algorithm for handling partitional clustering problems.

Automatic Segmentation of Liver from Abdominal Computed Tomography Images using Energy Feature

Liver Segmentation is one of the challenging tasks in detecting and classifying liver tumors from Computed Tomography (CT) images. The segmentation of hepatic organ is more intricate task, owing to the fact that it possesses a sizeable quantum of vascularization. This paper proposes an algorithm for automatic seed point selection using energy feature for use in level set algorithm for segmentation of liver region in CT scans. The effectiveness of the method can be determined when used in a model to classify the liver CT images as tumorous or not. This involves segmentation of the region of interest (ROI) from the segmented liver, extraction of the shape and texture features from the segmented ROI and classification of the ROIs as tumorous or not by using a classifier based on the extracted features. In this work, the proposed seed point selection technique has been used in level set algorithm for segmentation of liver region in CT scans and the ROIs have been extracted using Fuzzy C Means clustering (FCM) which is one of the algorithms to segment the images. The dataset used in this method has been collected from various repositories and scan centers. The outcome of this proposed segmentation model has reduced the area overlap error that could offer the intended accuracy and consistency. It gives better results when compared with other existing algorithms. Fast execution in short span of time is another advantage of this method which in turns helps the radiologist to ascertain the abnormalities instantly.

Full-automatic seed point selection and initialization for digital image correlation robust to large rotation and deformation

Accurate initial guess plays a key role to realize full-automatic digital image correlation (DIC) analysis, especially when large rotational deformation presents in deformed images. In this work, an efficient, robust and full-automatic initial guess approach combining the speeded-up robust features (SURF) algorithm and the reliability-guided displacement tracking (RGDT) strategy is proposed, which can not only automatically select and update seed point, but also can effectively deal with target images with large deformation and rotation. The scale- and rotation-invariant SURF algorithm can extract and match a certain number of feature points from two images even though the significant deformation and rotation present. The Euclidean distance and deformation information (including displacement and major orientation rotation angle) of the best-matched point are used to choose the seed point and determine its initial guess, respectively. Then, the RGDT strategy is then employed to continue the DIC analysis of rest calculation points. Compared with existing path-dependent initial guess using RGDT, the proposed method not only can automatically select seed points without manual intervention, but also can provide accurate initial value estimation for the deformed images in the presence of large rotation and/or deformation. Furthermore, it has evident efficiency advantages over existing path-independent initial guess methods based on SIFT and RANSAC. The robustness and effectiveness of the proposed method are validated by numerical simulation tests and real experiments.

TERRAIN-ADAPTIVE GROUND FILTERING OF AIRBORNE LIDAR DATA BASED ON SALIENCY-AWARE THIN PLATE SPLINE

Abstract. Ground filtering separates the ground and non-ground points from point clouds, which is the essential process for DEM generation, semantic segmentation, model reconstruction and so forth. Considering the topologically complex terrain environments, the segmentation results are prone to be disturbed dealing with steep slopes, buildings, bridges, cliffs, etc. from Airborne LiDAR point clouds. In this paper, a saliency-aware Thin-Plate-Spline (SATPS) interpolation method is proposed including two steps: saliency division and adaptive regularized TPS interpolation with relative variance coefficient. Firstly, the point clouds are indexed in 2D grids and segments are clustered step probing toward 8-adjacent scanning directions. Then, the saliency of each grid is calculated according to the elevation variance of adjacent segments towards each scanning direction. Subsequently, grids of high ground saliency are considered as candidates for seed point selection and then clustered by region growing. The TPS surface is interpolated for each cluster loosely fitting to the seed points involving an adaptive relative variance coefficient which is according to ground saliency and elevation deviation. And finally, the ground points are extracted around the TPS surface. Experimental results indicate that the proposed SATPS algorithm achieves better Type 1 accuracy and total accuracy than the state-of-the-art algorithms in scenes with complex terrain structures, which is practical to generate DEM products.

PLANAR POLYGONS DETECTION IN LIDAR SCANS BASED ON SENSOR TOPOLOGY ENHANCED RANSAC

Abstract. Detecting planar structures in point clouds is a very central step of the point cloud processing pipeline as many Lidar scans, in particular in anthropic environments, present such planar structures. Many improvements have been proposed to RANSAC and the Hough transform, the two major types of plane detection methods. An important limitation however is that these methods detect planes running across the whole scene instead of more localized planar patches. Moreover, they do not exploit the sensor information that often comes with Lidar point cloud (sensor topology and optical center position in particular). In this paper we address both issues: we aim at detecting planar polygons that have a limited spatial extent, and we exploit sensor topology. The latter is used to enhance a RANSAC framework on two aspects: to make seed points selection more local and to define more compact sets of inliers through sensor space region growing.

Image segmentation using fuzzy competitive learning based counter propagation network

Image segmentation is the method of partitioning an image into some homogenous regions that are more meaningful for its better understanding and examination. Soft computing methods having the capabilities of achieving artificial intelligence are predominately used to perform the task of segmentation. Due to the variability and the uncertainty present in natural scenes, segmentation is a complicated task to perform with the help of conventional image segmentation techniques. Therefore, in this article a hybrid Fuzzy Competitive Learning based Counter Propagation Network (FCPN) is proposed for the segmentation of natural scene images. This method compromises of the uncertainty handling capabilities of the fuzzy system and proficiency of parallel learning ability of neural network. To identify the number of clusters automatically in less computational time, the instar layer of Counter propagation network (CPN) has been trained by using Fuzzy competitive learning (FCL). The outstar layer of counter propagation network is trained by using Grossberg learning for obtaining the desired output. Region growing method having the tendency to correctly identify edges with simplicity is used for initial seed point selection. Then, the most similar regions in the image are clustered and the number of clusters is estimated automatically. Finally, by identifying the cluster centers the images are segmented. Bacterial foraging algorithm is used to initialize the initial weights to the network, which helps the proposed method in achieving low convergence ratio with higher accuracy. Results validated the higher performance of proposed FCPN method when compared with other states-of-the-art methods. For future work, some other adaptive methods like the fuzzy model-based network can be used to identify multiple object regions and classifying them among separate clusters.

Corrigendum to "A Novel Neutrosophic Method for Automatic Seed Point Selection in Thyroid Nodule Images".

[This corrects the article DOI: 10.1155/2019/7632308.].

A Novel Neutrosophic Method for Automatic Seed Point Selection in Thyroid Nodule Images.

The thyroid nodule is one of the endocrine issues caused by an irregular cell development. This rate of survival can be improved by earlier nodule detection. Accordingly, the accurate recognition of the nodule is of the utmost importance in providing powerful results in building the survival rate. The reduction in the accuracy of manual or semiautomatic segmentation methods for thyroid nodule detection is due to many factors, basically, the lack of experience of the sonographer and latency of operation. Most lesion regions in ultrasound images are homogeneous. Therefore, the value of entropy in these regions is high compared to its neighbours. Based on this criterion, a novel procedure for automatically selecting the seed point in thyroid nodule images is proposed. The proposed system consists of three components: neutrosophic image enhancement and speckle reduction to reduce speckle noise and automatic seed selection algorithm extracted from the centre of candidate block in ultrasound thyroid images based on the principle that most of its Higher Order Spectra Entropies (HOSE) from Radon Transform (RT) at different angles are within the range between average and maximum entropies, and the region growing image segmentation is applied with the constant threshold. The performance of proposed automatic segmentation method is compared with other methods in terms of calculating, True Positive (TP) value (96.44 ± 3.01%), False Positive (FP) value (3.55 ± 1.45%), Dice Coefficient (DC) value (92.24 ± 6.47%), Similarity Index (SI) (80.57 ± 1.06%), and Hausdroff Distance (HD) (0.42 ± 0.24 pixels). The proposed system can be considered as an added value to the malignancy diagnosis in thyroid nodule by an endocrinologist.

Seed selection algorithm through K-means on optimal number of clusters

Clustering is one of the important unsupervised learning in data mining to group the similar features. The growing point of the cluster is known as a seed. To select the appropriate seed of a cluster is an important criterion of any seed based clustering technique. The performance of seed based algorithms are dependent on initial cluster center selection and the optimal number of clusters in an unknown data set. Cluster quality and an optimal number of clusters are the important issues in cluster analysis. In this paper, the proposed seed point selection algorithm has been applied to 3 band image data and 2D discrete data. This algorithm selects the seed point using the concept of maximization of the joint probability of pixel intensities with the distance restriction criteria. The optimal number of clusters has been decided on the basis of the combination of seven different cluster validity indices. We have also compared the results of our proposed seed selection algorithm on an optimal number of clusters using K-Means clustering with other classical seed selection algorithms applied through K-Means Clustering in terms of seed generation time (SGT), cluster building Time (CBT), segmentation entropy and the number of iterations (NOTK−means). We have also made the analysis of CPU time and no. of iterations of our proposed seed selection method with other clustering algorithms.