Article

Snakes: Active Contour Models

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

A snake is an energy-minimizing spline guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges. Snakes are active contour models: they lock onto nearby edges, localizing them accurately. Scale-space continuation can be used to enlarge the capture region surrounding a feature. Snakes provide a unified account of a number of visual problems, including detection of edges, lines, and subjective contours, motion tracking, and stereo matching. The authors have used snakes successfully for interactive interpretation, in which user-imposed constraint forces guide the snake near features of interest.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Snake algorithm [7], particularly when integrated with deep learning (i.e., deep snake), presents a promising solution to these challenges. Unlike conventional semantic segmentation algorithms [8], which predict pixel-level semantic maps [9], the deep snake model generates initial object-level contours and refines them through vertex-wise offsets. ...
... Nevertheless, evolving contour points effectively to accurately fit object boundaries is challenging, with previous methods achieving limited success, particularly in medical imaging. Most existing approaches [7], [10]- [12] conceptualize the contour as a graph and employ graph convolutional networks to model snake evolution as a topological problem. While these frameworks offer a structured representation, they typically overlook the dynamic state-space transformation inherent in contour evolution, thereby limiting their effectiveness. ...
... Deep snake algorithms, which extend traditional active contour models [7] (ACMs) by incorporating deep learning techniques, demonstrate significant potential in multi-organ segmentation. By focusing on contour evolution rather than pixel-wise classification, these models are capable of generating smooth and realistic boundaries, even in scenarios with blurred edges or complex backgrounds. ...
Preprint
Multi-organ segmentation is a critical yet challenging task due to complex anatomical backgrounds, blurred boundaries, and diverse morphologies. This study introduces the Gradient-aware Adaptive Momentum Evolution Deep Snake (GAMED-Snake) model, which establishes a novel paradigm for contour-based segmentation by integrating gradient-based learning with adaptive momentum evolution mechanisms. The GAMED-Snake model incorporates three major innovations: First, the Distance Energy Map Prior (DEMP) generates a pixel-level force field that effectively attracts contour points towards the true boundaries, even in scenarios with complex backgrounds and blurred edges. Second, the Differential Convolution Inception Module (DCIM) precisely extracts comprehensive energy gradients, significantly enhancing segmentation accuracy. Third, the Adaptive Momentum Evolution Mechanism (AMEM) employs cross-attention to establish dynamic features across different iterations of evolution, enabling precise boundary alignment for diverse morphologies. Experimental results on four challenging multi-organ segmentation datasets demonstrate that GAMED-Snake improves the mDice metric by approximately 2% compared to state-of-the-art methods. Code will be available at https://github.com/SYSUzrc/GAMED-Snake.
... Deformable models utilize flexible curves to represent the shape of the WBCs being segmented. The active contour model, or snake model [16], has been widely applied for segmenting WBC nuclei from blood smear images in several studies [17][18][19][20]. Wenhua et al. [21] employed a level set method, followed by a Canny edge detector, to address the issue of unclear boundaries in segmentation. ...
... Additionally, it performs classification and regression outputs separately for multiple feature maps. The sizes of the candidate boxes are set to [16,32,64,128,256], with aspect ratios of [0.5, 1.0, 2.0]. ...
Preprint
Full-text available
The classification and statistics of white blood cells (WBCs) are critical steps in the microscopic examination of blood smears. Traditional manual microscopy methods are time-consuming and labor-intensive, while machine learning-based automated detection approaches require a substantial amount of labeled data for model training, leading to high costs. To address this issue, this paper proposes a two-stage semi-supervised deep learning method for WBC detection. In the first stage, a region proposal network (RPN) with ResNet50 as the backbone is employed for the localization and segmentation of white blood cell images. In the second stage, a semi-supervised learning framework is utilized to train the WBC classifier. The model is trained and tested using 1,510 labeled blood cell microscopy images with WBC localization boxes. The proposed semi-supervised model achieves a classification accuracy of 86%, which is 3.2% higher than that of the fully supervised model. Furthermore, this two-stage model is compared with two end-to-end models, FasterRCNN and RetinaNet. The results demonstrate that the proposed two-stage model achieves an accuracy of 83.7% and a recall of 85.1% in detection tasks, both exceeding those of the FasterRCNN and RetinaNet models. Compared to a one-stage WBC detection model, the two-stage detection method allows for more thorough training of the WBC classifier, thereby enhancing overall detection performance.
... Edge-based methods aim to encourage an evolving contour towards the edges in an image and normally require an edge detector function [8]. The first edge-based variational approach was devised by Kass et al. [22] with the famous snakes model, this was further developed by Casselles et al. [8] who introduced the Geodesic Active Contour (GAC) model. Region-based global segmentation models include the well known works of Mumford-Shah [29] and Chan-Vese [11]. ...
... Update u k to u k+1 using the AOS scheme (22). end for u * ← u k . ...
Preprint
Selective segmentation is an important application of image processing. In contrast to global segmentation in which all objects are segmented, selective segmentation is used to isolate specific objects in an image and is of particular interest in medical imaging -- permitting segmentation and review of a single organ. An important consideration is to minimise the amount of user input to obtain the segmentation; this differs from interactive segmentation in which more user input is allowed than selective segmentation. To achieve selection, we propose a selective segmentation model which uses the edge-weighted geodesic distance from a marker set as a penalty term. It is demonstrated that this edge-weighted geodesic penalty term improves on previous selective penalty terms. A convex formulation of the model is also presented, allowing arbitrary initialisation. It is shown that the proposed model is less parameter dependent and requires less user input than previous models. Further modifications are made to the edge-weighted geodesic distance term to ensure segmentation robustness to noise and blur. We can show that the overall Euler-Lagrange equation admits a unique viscosity solution. Numerical results show that the result is robust to user input and permits selective segmentations that are not possible with other models.
... Equation (26) can be used to compute the tangential component of Equation (27) for the problem (P1-P2). The normal component requires in addition the use of Proposition (12). Substituting the data of problem (P2) into equation (21), we obtain ...
... It has a fixed maximum possible value of C ≥ 90% of contour reached, as an example. Algorithm 1 Contour parametrization process 1: procedure 2: Initialize: 3: {ϕ j } k=0 ← points over initial curve 4: {(c i , y i )} ← data to define ρ 5: ∆t ← fixed step size 6: loop over k: 7: if there is (c i , y i ) not inside the curve then # for all ϕ k j compute: 11: Pix k j ← pixel value 12: ...
Preprint
Full-text available
We present a new implementation of anisotropic mean curvature flow for contour recognition. Our procedure couples the mean curvature flow of planar closed smooth curves, with an external field from a potential of point-wise charges. This coupling constrains the motion when the curve matches a picture placed as background. We include a stability criteria for our numerical approximation.
... Classical pioneering segmentation techniques such as active contour models (ACM) [9], k-means [10], and graphbased segmentation (GS) [11] [12] impose global and local data and geometry constraints on the masks. As a result, these techniques are sensitive to initialization and require heuristics such as point resampling, making them unsuitable for modern applications. ...
... The advent of convolutional neural networks (CNNs) and large annotated datasets has revolutionized image segmentation, surpassing traditional algorithms such as active contour models (ACM) [9], k-means [10], and graph-based segmentation (GS) [11]. Traditional methods often struggle with noise susceptibility, intricate boundary handling, and sensitivity to data quality [21,22,23]. ...
Article
Full-text available
Our work tackles the fundamental challenge of image segmentation in computer vision, which is crucial for diverse applications. While supervised methods demonstrate proficiency, their reliance on extensive pixel-level annotations limits scalability. We introduce DynaSeg, an innovative unsuper-vised image segmentation approach that overcomes the challenge of balancing feature similarity and spatial continuity without relying on extensive hyperparameter tuning. Unlike traditional methods, DynaSeg employs a dynamic weighting scheme that automates parameter tuning, adapts flexibly to image characteristics, and facilitates easy integration with other segmentation networks. By incorporating a Silhouette Score Phase, DynaSeg prevents undersegmentation failures where the number of predicted clusters might converge to one. DynaSeg uses CNN-based and pre-trained ResNet feature extraction, making it computationally efficient and more straightforward than other complex models. Experimental results showcase state-of-the-art performance, achieving a 12.2% and 14.12% mIOU improvement over current unsupervised segmentation approaches on COCO-All and COCO-Stuff datasets, respectively. We provide qualitative and quantitative results on five benchmark datasets, demonstrating the efficacy of the proposed approach.
... An active contours or snake is an energy-minimizing spline, guided by external constraint forces and influenced by image forces that pull it towards features such as lines and edges [14]. Active contours take an initial condition that is automatic or set up by the user. ...
Preprint
Animal biometrics is a challenging task. In the literature, many algorithms have been used, e.g. penguin chest recognition, elephant ears recognition and leopard stripes pattern recognition, but to use technology to a large extent in this area of research, still a lot of work has to be done. One important target in animal biometrics is to automate the segmentation process, so in this paper we propose a segmentation algorithm for extracting the spots of Diploglossus millepunctatus, an endangered lizard species. The automatic segmentation is achieved with a combination of preprocessing, active contours and morphology. The parameters of each stage of the segmentation algorithm are found using an optimization procedure, which is guided by the ground truth. The results show that automatic segmentation of spots is possible. A 78.37 % of correct segmentation in average is reached.
... The tracking method adopted here is a modified active contour model. The active contour, also called a snake, is a framework in computer vision for delineating an object outline from a possibly noisy 2D image (Kass et al. 1988;Pal & Pal 1993). A snake is an energy minimizing, deformable spline influenced by constraint and image forces that pull it towards object contours and internal forces that resist deformation. ...
Preprint
Full-text available
We design an event recognition-analysis system that can automatically detect solar type III radio burst and can mine information of the burst from the dynamic spectra observed by Nancay Decameter Array (NDA). We investigate the frequency drift rate of type III bursts and the speed of electron beams responsible for the generation of the bursts. Several computer vision methods are used in this automatic analysis system. The Hough transform is performed to recognize the line segment associated with type III bursts in the dynamic spectra. A modified active contour model is used to track the backbone of the burst and estimate the frequency drift rate at different frequency channels. We run this system on the NDA data from 2012 to 2017, and give a statistical survey of the event number distribution, the starting and stopping frequencies of bursts, the frequency dependence of the drift rate, and the exciter speed using three corona density models. The median value of the average frequency drift rates is about 6.94MHz/s for 1389 simple well-isolated type III bursts detected in the frequency range 10--80 MHz of NDA observation. The frequency drift rate changes with frequency as df/dt=0.0672f1.23df/dt = -0.0672 f^{1.23} from a least-squares fitting. The average exciter speed is about 0.2c based the density models. We do not find any significant dependence of the drift rate and the exciter speed on the solar activity cycle.
... Most algorithms use intensity thresholds to detect the worm's body and then use binary image operations to extract a centerline [14,15,16]. Here we use an open active contour approach [17,18] to extract the centerline from dark field images with modifications to account for cases when the worm's body crosses over itself as occurs during so-called "Omega Turns." In principle any method, automated or otherwise, that detects the centerlines should be sufficient. ...
Preprint
Advances in optical neuroimaging techniques now allow neural activity to be recorded with cellular resolution in awake and behaving animals. Brain motion in these recordings pose a unique challenge. The location of individual neurons must be tracked in 3D over time to accurately extract single neuron activity traces. Recordings from small invertebrates like C. elegans are especially challenging because they undergo very large brain motion and deformation during animal movement. Here we present an automated computer vision pipeline to reliably track populations of neurons with single neuron resolution in the brain of a freely moving C. elegans undergoing large motion and deformation. 3D volumetric fluorescent images of the animal's brain are straightened, aligned and registered, and the locations of neurons in the images are found via segmentation. Each neuron is then assigned an identity using a new time-independent machine-learning approach we call Neuron Registration Vector Encoding. In this approach, non-rigid point-set registration is used to match each segmented neuron in each volume with a set of reference volumes taken from throughout the recording. The way each neuron matches with the references defines a feature vector which is clustered to assign an identity to each neuron in each volume. Finally, thin-plate spline interpolation is used to correct errors in segmentation and check consistency of assigned identities. The Neuron Registration Vector Encoding approach proposed here is uniquely well suited for tracking neurons in brains undergoing large deformations. When applied to whole-brain calcium imaging recordings in freely moving C. elegans, this analysis pipeline located 150 neurons for the duration of an 8 minute recording and consistently found more neurons more quickly than manual or semi-automated approaches.
... Active Contours are energy-minimizing spline curves, guided by external constraint forces and influenced by image forces that pull it towards features such as lines and edges [15]. The internal forces in a spline curve impose smoothness constraints. ...
Preprint
Non-intrusive biometrics of animals using images allows to analyze phenotypic populations and individuals with patterns like stripes and spots without affecting the studied subjects. However, non-intrusive biometrics demand a well trained subject or the development of computer vision algorithms that ease the identification task. In this work, an analysis of classic segmentation approaches that require a supervised tuning of their parameters such as threshold, adaptive threshold, histogram equalization, and saturation correction is presented. In contrast, a general unsupervised algorithm using Markov Random Fields (MRF) for segmentation of spots patterns is proposed. Active contours are used to boost results using MRF output as seeds. As study subject the Diploglossus millepunctatus lizard is used. The proposed method achieved a maximum efficiency of 91.11%91.11\%.
... Active Contour Models (ACM): Kass, Witkin, and Terzopoulos [45] introduced Active Contour Models (ACM), or snakes, in 1988, revolutionizing image segmentation. Their method initiates a deformable contour near the object, iteratively adjusting it to minimize an energy function, balancing internal smoothness and external feature attraction forces. ...
Thesis
Full-text available
Automatic visual speech recognition (AVSR) techniques are increasingly prevalent in various domains, including manufacturing, public use, and multimedia devices, making Visual Speech Recognition (VSR) a promising technology that can improve communication accessibility for people with hearing impairments. However, most existing VSR systems are designed for languages like English, leaving a gap for languages like Arabic, which is spoken by over 400 million people worldwide and has unique linguistic and phonetic characteristics. This thesis presents a novel framework for Arabic Visual Speech Recognition (BlidAV10), which aims to address this gap and cater to the needs of the Arabic hearing impaired community. The framework integrates state-of-the-art deep learning techniques, such as Convolutional Neural Networks (CNNs) and Vision Transformers (ViT), to transcribe Arabic speech from visual cues accurately and efficiently. The framework also relies on a specialized Arabic dataset, which is carefully curated to capture the diversity and complexity of the Arabic language. This dataset serves as a benchmark for training and evaluating the VSR models, ensuring their robustness and reliability in real-world applications. The framework employs the deep learning techniques like YOLO, CNNs and ViT for robust mouth detection and recognition, which enables the extraction of crucial visual features for accurate speech transcription. The experimental results show that the proposed framework achieves promising performance in enhancing communication accessibility for Arabic speakers with hearing impairments. The framework also demonstrates its effectiveness in handling various linguistic and phonetic variations of the Arabic language, opening up new possibilities for wider applications in real-world scenarios. This research contributes significantly to advancing Arabic Visual Speech Recognition technology, enriching the VSR landscape and fostering greater inclusivity in communication for Arabic speakers.
... Early traditional techniques were developed for image segmentation like thresholding [2], histogram-based, region growing [3], watershed [4], clustering [5]. Then the technology advanced and some new methods were developed like sparsity-based [6], Markov random field [7], active contours [8] and graph cuts [9]. In recent years Deep Learning (DL) models have been developed with performance improvements and higher accuracy for image segmentation. ...
Article
Full-text available
Deep learning has played a vital role in advancing medical research, particularly in brain tumor segmentation. Despite using numerous deep learning algorithms for this purpose, accurately and reliably segmenting brain tumors remains a significant challenge. Segmentation of precise tumors is essential for the effective treatment of brain diseases. While deep learning offers a range of algorithms for segmentation, they still face limitations when analyzing medical images due to the variations in tumor shape, size, and location. This study proposes a deep learning approach combining a Generative Adversarial Network (GAN) with transfer learning and auto-encoder techniques to enhance brain tumor segmentation. The GAN incorporates a generator and discriminator to generate superior segmentation outcomes. In the generator, we applied downsampling and upsampling for tumor segmentation. In addition, an autoencoder is applied in which the encoder retains as much information as possible and then the decoder with those encodings reconstructs the image. At the bottleneck, the transfer learning technique is applied using the DenseNet model. By combining autoencoder techniques with transfer learning methodologies in GANs feature learning is enhanced, training time is reduced, and stability is increased. In this work, we enhanced the accuracy of brain tumor segmentation and even achieved better results for tumors having small size. We train and evaluate our proposed model using the publicly available BraTS 2021 dataset. The experimental result shows a dice score of 0.94 for the whole tumor, 0.86 for the tumor core, and 0.82 for the enhancing tumor. It is also shown that we achieve 2% to 4% higher accuracy as compared to other methods.
... Segmentation plays a vital role in a wide range of applications, including driverless vehicles, medical image analysis, augmented reality, and video surveillance, among others [9]. A multitude of image segmentation algorithms have been devised in the literature, ranging from basic techniques, e.g., histogram-based bundling, thresholding [10], k-means clustering [11], region-growing [12], and watersheds [13], to more sophisticated methods such as graph cuts [14], active contours [15], conditional and Markov random fields [16], and sparsity-based approaches [17,18]. Currently acknowledged as the next generation of image segmentation algorithms, Deep Learning (DL) models exhibit significant performance gains in recent years. ...
Preprint
Full-text available
Segment Anything Model 2 (SAM 2) is a state-of-the-art development by Meta AI Research, designed to address the limitations of its predecessor, SAM, particularly in the realm of video segmentation. SAM 2 employs a transformer-based architecture enhanced with streaming memory, enabling real-time processing for both images and videos. This advancement is important given the exponential growth of multimedia content and the subsequent demand for efficient video analysis. Utilizing the SA-V dataset, SAM 2 excels in handling the intricate spatio-temporal dynamics inherent in video data, ensuring accurate and efficient segmentation. Key features of SAM 2 include its ability to provide real-time segmentation with minimal user interaction, maintaining robust performance even in dynamic and cluttered visual environments. This study provides a comprehensive overview of SAM 2, detailing its architecture, functionality, and diverse applications. It further explores the model's potential in improving practical implementations across various domains, emphasizing its significance in advancing real-time video analysis.
... Image segmentation serves as a critical step in the realm of image analysis and has garnered significant utilization across numerous pertinent domains. Image segmentation can be broadly categorized into four major classes: region-based segmentation methods, comprising seed region growing and watershed algorithms; energy functional methods, including Snake models [3] and their derivatives; edgebased segmentation, including Roberts gradient operator [4], Sobel gradient operator, etc; and thresholding methods, such as the application of Two-dimensional(2D)OTSU in moon rocks and craters segmentation [5]. ...
Article
Full-text available
With the development of society and changes in the human living environment, people are increasingly attaching importance to their own health. Regarding medical imaging examinations of certain parts of the body, the process of medical image segmentation has become extremely important. This paper presents a novel hybrid algorithm: SAOBL-IA, a fusion of the Simulated Annealing(SA), Opposition-based Learning(OBL)and Island Algorithm(IA). The Island Algorithm itself suffers from slow convergence speed and the tendency to get stuck in local optimum. To address these limitations, we introduce opposition-based learning to enhance the search range and avoid local optimum. Furthermore, we leverage the simulated annealing approach to accelerate the convergence of SAOBL-IA. Comparing the experimental results, it can be seen that SAOBL-IA has better comprehensive performance. Subsequently, the SAOBL-IA algorithm is utilized in conjunction with an optimized two-dimensional OTSU fusion segmentation technique for the purpose of medical image processing. This study proposes an application of image segmentation based on the SAOBL-IA. The segmentation of pixels around the background and target regions using the two-dimensional OTSU method faces challenges in terms of accuracy. To address this issue, an adaptive thresholding technique known as Adaptive Forking is employed for optimization. By determining the slope of the fork based on the misclassified pixel ratio, enhanced segmentation accuracy can be achieved. This improved approach is then integrated with the SAOBL-IA algorithm and applied to the segmentation of lung medical images. The experimental findings show that the amalgamation of SAOBL-IA with the adaptive two-dimensional OTSU segmentation approach, as proposed in this study, manifests superior segmentation speed and enhanced precision in the context of medical image segmentation.
... Contour tracing is a major contributor to the efficiency of the feature extraction process of silhouettes, for example to implement Freeman's chain-coded Curves, [FD77,LW11]. The most common contour tracing algorithms are the Square Tracing [Pra01], Moore-Neighbor Tracing [SHB08], the Radial Sweep [RAB12], Theo Pavlids' algorithm [Pav82], Snakes algorithm [KWT88], Amoeba algorithm [IV00], Topological-hierarchical algorithms [KC14]; among others so called Fast Contour-Tracing algorithms [BD18,SCS + 16]. ...
Conference Paper
Full-text available
Silhouette extraction involves separating objects of interest from their background, which has several applications in image processing. Among the silhouette extraction techniques, contour tracing is commonly applied to images with a uniform background. This paper introduces a novel contribution to contour tracing techniques, utilizing the Wall-Follower Algorithm (WFA) to extract silhouettes with uniform backgrounds, or binary images. The algorithm is based on the analogy of a follower sequentially walking aside the external boundary of a wall, without separating a hand from it; then, the follower walks tagging silhouette pixels along the way until returning to the initial position and direction. Experimentation on vehicle technical drawings, satellite views of bodies of water and photographs of plants shows its effectiveness in producing high-quality silhouettes while showing some advantages over existing techniques. They include quickness in obtaining a solution, efficiency and ability to handle complex contours, and the option to simplify the results by reducing the percentage of saved points that trace the perimeter, based on object characteristics. The robustness of the algorithm suggests it as a promising alternative with diverse applications in image analysis, computer-aided design, and 3D object reconstruction, by extruding silhouettes, the latter being the main motivation for this contribution.
... Active contour or deformable models provide a practical framework for object segmentation and have been widely used in image segmentation. Kass et al. [25] first introduced the active contour method, evolving a contour towards object boundaries by minimizing an energy function. Osher et al. [35] and Sethian et al. [1] developed the level set method, using level sets of a higher dimensional function to enable the implicit representation of curves. ...
Preprint
Full-text available
This paper presents a novel shape prior segmentation method guided by the Harmonic Beltrami Signature (HBS). The HBS is a shape representation fully capturing 2D simply connected shapes, exhibiting resilience against perturbations and invariance to translation, rotation, and scaling. The proposed method integrates the HBS within a quasi-conformal topology preserving segmentation framework, leveraging shape prior knowledge to significantly enhance segmentation performance, especially for low-quality or occluded images. The key innovation lies in the bifurcation of the optimization process into two iterative stages: 1) The computation of a quasi-conformal deformation map, which transforms the unit disk into the targeted segmentation area, driven by image data and other regularization terms; 2) The subsequent refinement of this map is contingent upon minimizing the L2L_2 distance between its Beltrami coefficient and the reference HBS. This shape-constrained refinement ensures that the segmentation adheres to the reference shape(s) by exploiting the inherent invariance, robustness, and discerning shape discriminative capabilities afforded by the HBS. Extensive experiments on synthetic and real-world images validate the method's ability to improve segmentation accuracy over baselines, eliminate preprocessing requirements, resist noise corruption, and flexibly acquire and apply shape priors. Overall, the HBS segmentation framework offers an efficient strategy to robustly incorporate the shape prior knowledge, thereby advancing critical low-level vision tasks.
Article
Full-text available
Plant phenotyping plays a crucial role in crop science and plant breeding. However, traditional methods often involve time-consuming and manual observations. Therefore, it is essential to develop automated, sensor-driven techniques that can provide objective and rapid information. Various methods rely on camera systems, including RGB, multi-spectral, and hyper-spectral cameras, which offer valuable insights into plant physiology. In recent years, 3D sensing systems such as laser scanners have gained popularity due to their ability to capture structural plant parameters that are difficult to obtain using spectral sensors. Unlike images, point clouds are not structured and require pre-processing steps to extract precise information and handle noise or missing points. One approach is to generate mesh-based surface representations using triangulation. A key challenge in the 3D surface reconstruction of plants is the pre-processing of point clouds, which involves removing non-plant noise from the scene, segmenting point clouds from populations to individual plants, and further dividing individual plants into their respective organs. In this study, we will not focus on the segmentation aspect but rather on the other pre-processing steps, like denoising parameters, which depend on the data type. We present an automated pipeline for converting high-resolution point clouds into surface models of plants. The pipeline incorporates additional pre-processing steps such as outlier removal, denoising, and subsampling to ensure the accuracy and quality of the reconstructed surfaces. Data were collected using three different sensors: a handheld scanner, a terrestrial laser scanner (TLS), and a mobile mapping platform, under varying conditions from controlled laboratory environments to complex field settings. The investigation includes five different plant species, each with distinct characteristics, to demonstrate the potential of the pipeline. In a next step, phenotypic traits such as leaf area, leaf area index (LAI), and leaf angle distribution (LAD) were calculated to further illustrate the pipeline’s potential and effectiveness. The pipeline is based on the Open3D framework and is available open source.
Article
Full-text available
This paper introduces a new method for the region segmentation of images. The approach is based on the raster-scan paradigm and builds the segments incrementally. The pixels are processed in the raster-scan order, while the construction of the segments is based on a distance metric in regard to the already segmented pixels in the neighbourhood. The segmentation procedure operates in linear time according to the total number of pixels. The proposed method, named the RSM (raster-scan segmentation method), was tested on selected images from the popular benchmark datasets MS COCO and DIV2K. The experimental results indicate that our method successfully extracts regions with similar pixel values. Furthermore, a comparison with two of the well-known segmentation methods—Watershed and DBSCAN—demonstrates that the proposed approach is superior in regard to efficiency while yielding visually similar results.
Article
Full-text available
Automatic large-scale building extraction from the LiDAR point clouds and remote sensing images is a growing focus in the fields of the sensor applications and remote sensing. However, this building extraction task remains highly challenging due to the complexity of building sizes, shapes, and surrounding environments. In addition, the discreteness, sparsity, and irregular distribution of point clouds, lighting, and shadows, as well as occlusions of the images, also seriously affect the accuracy of building extraction. To address the above issues, we propose a new unsupervised building extraction algorithm PBEA (Point and Pixel Building Extraction Algorithm) based on a new dual P-snake model (Dual Point and Pixel Snake Model). The proposed dual P-snake model is an enhanced active boundary model, which uses both point clouds and images simultaneously to obtain the inner and outer boundaries. The proposed dual P-snake model enables interaction and convergence between the inner and outer boundaries to improve the performance of building boundary detection, especially in complex scenes. Using the dual P-snake model and polygonization, this proposed PBEA can accurately extract large-scale buildings. We evaluated our PBEA and dual P-snake model on the ISPRS Vaihingen dataset and the Toronto dataset. The experimental results show that our PBEA achieves an area-based quality evaluation metric of 90.0% on the Vaihingen dataset and achieves the area-based quality evaluation metric of 92.4% on the Toronto dataset. Compared with other methods, our method demonstrates satisfactory performance.
Article
Full-text available
As semiconductor manufacturing processes continue to witness increased integration density and design complexity, semiconductor wafers are experiencing a growing diversity and complexity of defects. While previous research in wafer map classification using deep learning has made significant advancements in dealing with single defect patterns, the classification of mixed-type defects has received less attention due to their considerably higher difficulty level compared to single defects. This research addresses this critical gap, emphasizing the need for improved methods to classify mixed-type defects, which are more complex and challenging. To tackle this challenge, this paper introduces the active contour-based lightweight depthwise network (AC-LDN) model for the classification of multi-defect wafer map patterns. Initially, multi-defect features are extracted using an active contour-based segmentation model. Subsequently, the learning model employs a depthwise convolutional neural network (CNN) architecture that combines separable CNN and dilated CNN techniques. This unique approach optimizes the model in the separable segment while effectively addressing defect complexity in the depthwise segments. Consequently, AC-LDN outperforms other state-of-the-art models, offering a balance between lightweight characteristics and high accuracy. The proposed method demonstrates its superiority over previous models when evaluated on the extsdsensive multi-wafer map dataset, achieving an average classification accuracy exceeding 98% and a confusion matrix coefficient surpassing 0.97.
Article
Full-text available
The acurate segmentation and classification of nuclei in histological images are crucial for the diagnosis and treatment of colorectal cancer. However, the aggregation of nuclei and intra-class variability in histology images present significant challenges for nuclei segmentation and classification. In addition, the imbalance of various nuclei classes exacerbates the difficulty of nuclei classification and segmentation using deep learning models. To address these challenges, we present a novel attention-enhanced residual refinement network (AER-Net), which consists of one encoder and three decoder branches that have same network structure. In addition to the nuclei instance segmentation branch and nuclei classification branch, one branch is used to predict the vertical and horizontal distance from each pixel to its nuclear center, which is combined with output by the segmentation branch to improve the final segmentation results. The AER-Net utilizes an attention-enhanced encoder module to focus on more valuable features. To further refine predictions and achieve more accurate results, an attention-enhancing residual refinement module is employed at the end of each encoder branch. Moreover, the coarse predictions and refined predictions are combined by using a loss function that employs cross-entropy loss and generalized dice loss to efficiently tackle the challenge of class imbalance among nuclei in histology images. Compared with other state-of-the-art methods on two colorectal cancer datasets and a pan-cancer dataset, AER-Net demonstrates outstanding performance, validating its effectiveness in nuclear segmentation and classification.
Article
Full-text available
Synopsis Artificial intelligence (AI) is poised to revolutionize many aspects of science, including the study of evolutionary morphology. While classical AI methods such as principal component analysis and cluster analysis have been commonplace in the study of evolutionary morphology for decades, recent years have seen increasing application of deep learning to ecology and evolutionary biology. As digitized specimen databases become increasingly prevalent and openly available, AI is offering vast new potential to circumvent long-standing barriers to rapid, big data analysis of phenotypes. Here, we review the current state of AI methods available for the study of evolutionary morphology, which are most developed in the area of data acquisition and processing. We introduce the main available AI techniques, categorizing them into 3 stages based on their order of appearance: (1) machine learning, (2) deep learning, and (3) the most recent advancements in large-scale models and multimodal learning. Next, we present case studies of existing approaches using AI for evolutionary morphology, including image capture and segmentation, feature recognition, morphometrics, and phylogenetics. We then discuss the prospectus for near-term advances in specific areas of inquiry within this field, including the potential of new AI methods that have not yet been applied to the study of morphological evolution. In particular, we note key areas where AI remains underutilized and could be used to enhance studies of evolutionary morphology. This combination of current methods and potential developments has the capacity to transform the evolutionary analysis of the organismal phenotype into evolutionary phenomics, leading to an era of “big data” that aligns the study of phenotypes with genomics and other areas of bioinformatics.
Article
Full-text available
This paper reports two distinct vortex-flame interaction mechanisms existing in a self-excited V-shaped premixed flame. Time-resolved simultaneous measurements of particle image velocimetry (PIV) and OH* chem-iluminescence are employed to capture the mean and coherent flow and flame structures. The qualitative study demonstrates the importance of the outer vortex rings (OVR) in governing the flame front dynamics of two representative lean and near-stoichiometric cases. The interaction of the flame with the OVRs is further analyzed quantitatively during one OVR's entire lifespan. Results suggest that the vortex-flame interaction in the lean case is a weak one as the response of the flame's HRR to OVR's growth is interrupted by the tip extinction. This can be understood as a response to the flow perturbation excited in the Helmholtz mode of the plenum. However, the near-stoichiometric case yields a strong interaction that the peak HRR is synchronized with the OVR's peak circulation, which could contribute to a stronger thermoacoustic coupling that leads to the stronger pressure oscillation and frequency drift away from the Helmholtz mode.
Article
Full-text available
Both pathology and cytopathology still rely on recognizing microscopical morphologic features, and image analysis plays a crucial role, enabling the identification, categorization, and characterization of different tissue types, cell populations, and disease states within microscopic images. Historically, manual methods have been the primary approach, relying on expert knowledge and experience of pathologists to interpret microscopic tissue samples. Early image analysis methods were often constrained by computational power and the complexity of biological samples. The advent of computers and digital imaging technologies challenged the exclusivity of human eye vision and brain computational skills, transforming the diagnostic process in these fields. The increasing digitization of pathological images has led to the application of more objective and efficient computer-aided analysis techniques. Significant advancements were brought about by the integration of digital pathology, machine learning, and advanced imaging technologies. The continuous progress in machine learning and the increasing availability of digital pathology data offer exciting opportunities for the future. Furthermore, artificial intelligence has revolutionized this field, enabling predictive models that assist in diagnostic decision making. The future of pathology and cytopathology is predicted to be marked by advancements in computer-aided image analysis. The future of image analysis is promising, and the increasing availability of digital pathology data will invariably lead to enhanced diagnostic accuracy and improved prognostic predictions that shape personalized treatment strategies, ultimately leading to better patient outcomes.
Article
Full-text available
Ship contour extraction is vital for extracting the geometric features of ships, providing comprehensive information essential for ship recognition. The main factors affecting the contour extraction performance are speckle noise and amplitude inhomogeneity, which can lead to over-segmentation and missed detection of ship edges. Polarimetric synthetic aperture radar (PolSAR) images contain rich target scattering information. Under different transmitting and receiving polarization, the amplitude and phase of pixels can be different, which provides the potential to meet the uniform requirement. This paper proposes a novel ship contour extraction framework from PolSAR images based on polarization modulation. Firstly, the image is partitioned into the foreground and background using a super-pixel unsupervised clustering approach. Subsequently, an optimization criterion for target amplitude modulation to achieve uniformity is designed. Finally, the ship’s contour is extracted from the optimized image using an edge-detection operator and an adaptive edge extraction algorithm. Based on the contour, the geometric features of ships are extracted. Moreover, a PolSAR ship contour extraction dataset is established using Gaofen-3 PolSAR images, combined with expert knowledge and automatic identification system (AIS) data. With this dataset, we compare the accuracy of contour extraction and geometric features with state-of-the-art methods. The average errors of extracted length and width are reduced to 20.09 m and 8.96 m. The results demonstrate that the proposed method performs well in both accuracy and precision.
Article
Full-text available
Heart disease is a leading cause of mortality, with calcific aortic valve disease (CAVD) being the most prevalent subset. Being able to predict this disease in its early stages is important for monitoring patients before they need aortic valve replacement surgery. Thus, this study explored hydrodynamic, mechanical, and hemodynamic differences in healthy and very mildly calcified porcine small intestinal submucosa (PSIS) bioscaffold valves to determine any notable parameters between groups that could, possibly, be used for disease tracking purposes. Three valve groups were tested: raw PSIS as a control and two calcified groups that were seeded with human valvular interstitial and endothelial cells (VICs/VECs) and cultivated in calcifying media. These two calcified groups were cultured in either static or bioreactor-induced oscillatory flow conditions. Hydrodynamic assessments showed metrics were below thresholds associated for even mild calcification. Young’s modulus, however, was significantly higher in calcified valves when compared to raw PSIS, indicating the morphological changes to the tissue structure. Fluid–structure interaction (FSI) simulations agreed well with hydrodynamic results and, most notably, showed a significant increase in time-averaged wall shear stress (TAWSS) between raw and calcified groups. We conclude that tracking hemodynamics may be a viable biomarker for early-stage CAVD tracking.
Article
Full-text available
Machine vision is a desirable non-contact measurement method for hot forgings, as image segmentation has been a challenging issue in performance and robustness resulting from the diversity of working conditions for hot forgings. Thus, this paper proposes an efficient and robust active contour model and corresponding image segmentation approach for forging images, by which verification experiments are conducted to prove the performance of the segmentation method by measuring geometric parameters for forging parts. Specifically, three types of continuity parameters are defined based on the geometric continuity of equivalent grayscale surfaces for forging images; hence, a new image force and external energy functional are proposed to form a new active contour model, Geometric Continuity Snakes (GC Snakes), which is more percipient to the grayscale distribution characteristics of forging images to improve the convergence for active contour robustly; additionally, a generating strategy for initial control points for GC Snakes is proposed to compose an efficient and robust image segmentation approach. The experimental results show that the proposed GC Snakes has better segmentation performance compared with existing active contour models for forging images of different temperatures and sizes, which provides better performance and efficiency in geometric parameter measurement for hot forgings. The maximum positioning and dimension errors by GC Snakes are 0.5525 mm and 0.3868 mm, respectively, compared with errors of 0.7873 mm and 0.6868 mm by the Snakes model.
Article
Full-text available
Accurate medical image segmentation can assist doctors in observing lesion areas and making precise judgments. Effectively utilizing important multi-scale semantic information in local and global contexts is key to improving segmentation accuracy. In this paper, we present a multi-scale dual attention network (MSDA-Net), which enhances feature representation under different receptive fields and effectively utilizes the important multi-scale semantic information from both local and global contexts in medical images. MSDA-Net is a typical encoder–decoder structure and introduces a multi-receptive field densely connected module (MRD) in the decoder. This module captures semantic information across various receptive fields and utilizes dense connections to provide comprehensive and detailed semantic representations. Furthermore, a parallel dual-branch attention module (PDA), incorporating spatial and channel attention, focuses intensively on detailed features within lesion areas. This module enhances feature representation, facilitates the identification of disease boundaries, and improves the accuracy of segmentation. To validate the effectiveness of MSDA-Net, we conducted performance analyses on the CVC-ClinicDB, 2018 Data Science Bowl, ISIC 2018, and colon cancer slice datasets. We also compared our method with U-Net, UNet++, and other methods. The experimental results unequivocally demonstrate that MSDA-Net outperforms these methods, showcasing its superior performance in medical image segmentation tasks.
ResearchGate has not been able to resolve any references for this publication.