Region Segmentation Algorithm Research Articles

Automatic reorientation to generate short-axis myocardial PET images

BackgroundAccurately redirecting reconstructed Positron emission tomography (PET) images into short-axis (SA) images shows great significance for subsequent clinical diagnosis. We developed a system for automatic redirection and quantitative analysis of myocardial PET images.MethodsA total of 128 patients were enrolled for 18 F-FDG PET/CT myocardial metabolic images (MMIs), including 3 image classifications: without defects, with defects, and excess uptake. The automatic reorientation system includes five modules: regional division, myocardial segmentation, ellipsoid fitting, image rotation and quantitative analysis. First, the left ventricular geometry-based canny edge detection (LVG-CED) was developed and compared with the other 5 common region segmentation algorithms, the optimized partitioning was determined based on partition success rate. Then, 9 myocardial segmentation methods and 4 ellipsoid fitting methods were combined to derive 36 cross combinations for diagnostic performance in terms of Pearson correlation coefficient (PCC), Kendall correlation coefficient (KCC), Spearman correlation coefficient (SCC), and determination coefficient. Finally, the deflection angles were computed by ellipsoid fitting and the SA images were derived by affine transformation. Furthermore, the polar maps were used for quantitative analysis of SA images, and the redirection effects of 3 different image classifications were analyzed using correlation coefficients.ResultsOn the dataset, LVG-CED outperformed other methods in the regional division module with a 100% success rate. In 36 cross combinations, PSO-FCM and LLS-SVD performed the best in terms of correlation coefficient. The linear results indicate that our algorithm (LVG-CED, PSO-FCM, and LLS-SVD) has good consistency with the reference manual method. In quantitative analysis, the similarities between our method and the reference manual method were higher than 96% at 17 segments. Moreover, our method demonstrated excellent performance in all 3 image classifications.ConclusionOur algorithm system could realize accurate automatic reorientation and quantitative analysis of PET MMIs, which is also effective for images suffering from interference.

Abstract PO5-07-02: AI-assisted interpretation of PD-L1 CPS improves the precision medicine in Triple-negative breast cancer

Abstract Objective: The IMpassion 031 and KEYNOTE-355 clinical trial demonstrated the advantages of atezolizumab and pembrolizumab in the treatment of Triple-negative breast cancer (TNBC). However, the application of CPS in assessing PD-L1 expression is challenging. Pathologists have great variability in interpretation of PD-L1 CPS, the daily diagnostic workload is too heavy and their diagnostic efficiency is low, and they still have deficiencies in terms of accurate enumeration. It is of necessity to establish an objective and effective method which is highly repeatable. Methods: In this study, we established a deep learning-based artificial intelligence-assisted (AI-assisted) model which using cell detection and region segmentation algorithm. Three rounds of ring studies (RSs) were conducted. 12 pathologists of different level evaluate the CPS of PD-L1 (DAKO 22C3) in TNBC patients by visual assessment and AI-assisted model. Compare the difference, consistency and accuracy of the interpretation. Our research evaluated PD-L1 CPS expression with continuous score. Shapiro-Wilk (S-W) method performs normality test, difference analysis using Friedman M and Bonferroni calibration tests, consistency analysis is studied by intraclass correlation coefficient (ICC) and heat maps and box plots. Results: In the visual assessment, the interpretation results of PD-L1 (DAKO 22C3) CPS in different level pathologists have significant differences (P < 0.05), and the consistency of all the pathologists interpretation results was weak. Due to this strong inconsistency, it may result in a number of patients losing the opportunity to use ICIs for treatment. Moreover, the internal consistency of all pathologists in the visual assessment is moderate, in which the repeatability of junior pathologists is the worst, the ICC value is 0.664 (95%CI: 0.564-0.762). Through AI-assisted interpretation, there is no significant difference between all pathologists (P = 0.425), and the ICC value increased to 0.883 (95%CI:0.836-0.922) which improved the consistency of the interpretation results. In addition, through AI-assisted interpretation, the repeatability and accuracy of the interpretation results has been further upgraded. At the same time, the acceptance of AI results by junior pathologists are high, and 80% of the AI results are accepted. Conclusions: With the help of the AI-assisted diagnostic model, different levels of all the pathologists have achieved excellent consistency and repeatability in the interpretation of PD-L1 (DAKO 22C3) CPS. Moreover, the level of interpretation has been sought the rapid enhancement. It can be seen that AI-assisted diagnostic model provides a good approach to strengthen the consistency and repeatability in clinical practice. Our research also shows that PD-L1 CPS can receive precise interpretation from pathologists, so more patients have the opportunity to use ICIs. Therefore, these patients can benefit from treatment and achieve better survival. The interpretation results and system diagram of PD-L1 CPS by AI-assisted model. The consistency and repeatability of the interpretation results. (A)-(C) The consistency of the PD-L1 CPS interpretation among the different level of pathologists in three ring studies. (D)The repeatability of the interpretation results between 12 pathologists in RS1 and RS2. Accuracies and acceptance rates of different experienced pathologists. (A) Boxplots of scoring accuracies for pathologists in different levels. (B) Acceptance rates of all pathologists in total 50 slides. (C) Percent stacked column chart of different levels of pathologists with AI score acceptance. Citation Format: Jinze Li, Xinran Wang, Yueping Liu. AI-assisted interpretation of PD-L1 CPS improves the precision medicine in Triple-negative breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-07-02.

The Extraction of Roof Feature Lines of Traditional Chinese Village Buildings Based on UAV Dense Matching Point Clouds

Traditional Chinese buildings serve as a carrier for the inheritance of traditional culture and national characteristics. In the context of rural revitalization, achieving the 3D reconstruction of traditional village buildings is a crucial technical approach to promoting rural planning, improving living environments, and establishing digital villages. However, traditional algorithms primarily target urban buildings, exhibiting limited adaptability and less ideal feature extraction performance for traditional residential buildings. As a result, guaranteeing the accuracy and reliability of 3D models for different types of traditional buildings remains challenging. In this paper, taking Jingping Village in Western Hunan as an example, we propose a method that combines multiple algorithms based on the slope segmentation of the roof to extract feature lines. Firstly, the VDVI and CSF algorithms are used to extract the building and roof point clouds based on the MVS point cloud. Secondly, according to roof features, village buildings are classified, and a 3D roof point cloud is projected into 2D regular grid data. Finally, the roof slope is segmented via slope direction, and internal and external feature lines are obtained after refinement through Canny edge detection and Hough straight line detection. The results indicate that the CSF algorithm can effectively extract the roofs of I-shaped, L-shaped, and U-shaped traditional buildings. The accuracy of roof surface segmentation based on slope exceeds 99.6%, which is significantly better than the RANSAC algorithm and the region segmentation algorithm. This method is capable of efficiently extracting the characteristic lines of roofs in low-rise buildings within traditional villages. It provides a reference method for achieving the high-precision modeling of traditional village architecture at a low cost and with high efficiency.

MSegResRF-SPECT: A Novel Joint Classification Model of Whole Body Bone Scan Images for Bone Metastasis Diagnosis.

Whole-body bone scanning is a nuclear medicine technique with high sensitivity used for the diagnosis of bone-related diseases [e.g., bone metastases] that can be obtained by positron emission tomography[PET] or single-photon emission computed tomography[SPECT] imaging, depending on the different radiopharmaceuticals used. In contrast to the high sensitivity of the bone scan, it has low specificity, which leads to misinterpretation, causing adverse effects of unwarranted intervention or interruption to timely treatment. To address this problem, this paper proposes a joint model called mSegResRF-SPECT, which accomplishes for the first time the task of classifying whole-body bone scan images on a public SPECT dataset [BS-80K] for the diagnosis of bone metastases. The mSegResRF-SPECT adopts a multi-bone region segmentation algorithm to segment the whole body image into 13 regions, ResNet34 as an extractor to extract the regional features, and a random forest algorithm as a classifier. The experimental results of the proposed model show that the average accuracy, sensitivity, and F1 score of the model on the BS-80K dataset reached SOTA. The proposed method presents a promising solution for better bone scan classification methods.

Research on Human Lung Impedance Tomography Based on Soft Thresholding Image Segmentation and Reduced-Order Tikhonov Regularization

The lung is one of the most vital organs in the human body, and its condition is closely correlated with overall health. Electrical impedance tomography (EIT), as a biomedical imaging technique, often produces low-quality reconstructed images due to its inherent ill-posedness in solving the inverse problem. To address this issue, this paper proposes a soft-threshold region segmentation algorithm with a relaxation factor. This algorithm segments the reconstructed lung images into internal regions, edge regions, and background regions, resulting in clearer boundaries in the reconstructed images. This facilitates the intuitive identification of regions of interest by healthcare professionals. Additionally, this segmentation algorithm is suitably combined with a dimension-reduced Tikhonov regularization algorithm. By utilizing the joint capabilities of these algorithms, the partition points belonging to the background region can be excluded from the sought grayscale vector, thereby improving the ill-posedness of the image reconstruction process and enhancing the quality of image reconstruction. Finally, a 16-electrode human lung EIT simulation model is established for the thoracic region and verified through simulation. Experimental validation is conducted using a human lung tank simulation platform to further demonstrate the effectiveness of the proposed method.

Artificial Intelligence based real-time automatic detection and classification of skin lesion in dermoscopic samples using DenseNet-169 architecture

The most common challenge faced by dermoscopy images is the automatic detection of lesion features. All the existing solutions focus on complex algorithms to provide accurate detections. In this research work, proposed Online Tigerclaw Fuzzy Region Segmentation with Deep Learning Classification model, an intellectual model is proposed that provides discrimination of features with classification even in fine-grained samples. This model works on four different stages, which include the Boosted Anisotropic Diffusion filter with Recursive Pixel Histogram Equalization (BADF-RPHE) in the preprocessing stage. The next step is the proposed Online Tigerclaw Fuzzy Region Segmentation (OTFRS) algorithm for lesion area segmentation of dermoscopic images, which can achieve 98.9% and 97.4% accuracy for benign and malignant lesions, respectively. In the proposed OTFRS, an accuracy improvement of 1.4% is achieved when compared with previous methods. Finally, the increased robustness of lesion classification is achieved using Deep Learning Classification –DenseNet 169 with 500 images. The proposed approach was evaluated with accuracy classifications of 100% and 98.86% for benign and malignant lesions, respectively, and a processing time of less than 18 sec. In the proposed DensetNet-169 classification technique, an accuracy improvement of 3% is achieved when compared with other state-of-art methods. A higher range of true positive values is obtained for the Region of Convergence (ROC) curve, which indicates that the proposed work ensures better performance in clinical diagnosis for accurate feature visualization analysis. The methodology has been validated to prove its effectiveness and throw light on the lives of affected patients so they can resume normalcy and live long. The research work was tested in real-time clinical samples, which delivered promising and encouraging results in skin cell detection procedures.

Efficient railway track region segmentation algorithm based on lightweight neural network and cross-fusion decoder

To segment railway track regions in real-time for intrusion detection and improving security, this paper proposes an efficient railway track region segmentation network (ERTNet) based on the encoder–decoder architecture. Firstly, to ensure the lightweight of the encoder, depthwise convolution and the channel shuffle are utilized to construct sandglass-type feature extraction unit. Secondly, a feature-matching-based cross-fusion decoder is utilized to fuse deep and shallow feature maps. Thirdly, the knowledge distillation is employed with large-scale Deeplab v3+ as the teacher model to improve performance. Additionally, a loss function is proposed to penalize pixel points with large offsets. Finally, the ERTNet is validated on the self-built dataset, achieving an MIoU (Mean Intersection over Union) of 92.4% , which is 5.22% improvement over the benchmark model. ERTNet achieves a balance between segmentation accuracy and computational efficiency, requiring only 0.5 M parameters and 0.92 G FLOPs (Floating Point Operations).

Objective quantification of posterior capsule opacification after cataract surgery with swept-source optical coherence tomography

PurposeTo evaluate the application of swept-source optical coherence tomography (SS-OCT) and pentacam scheimpflug tomography in posterior capsule opacification (PCO) severity assessment.MethodsThe posterior capsule image region segmentation and adaptive threshold algorithm are used to process the SS-OCT scanned image to obtain the posterior capsule thickness (PCT). Scheimpflug tomography reconstructed and analysized by image J software can obtain the average gray value and evaluate the effectiveness with the two methods.ResultOne hundred sixty-two IOL eyes of 101 patients were divided into two groups, laser group (65 eyes) with the mean PCT was 8.0 ± 2.7 pixel unit and the mean gray value of the eyes was 66 ± 33 pixel unit. However, these figures in the control group (97 eyes) were 5.0 ± 0.9 and 11 ± 17. The sensitivity, specificity and area under curve(AUC) of SS-OCT PCT were 85%, 74% and 0.942,the sensitivity, specificity and AUC of Pentacam gray value were 91%, 76% and 0.947, respectively. After using the multivariable model of generalized estimation equation to corrected the dependence of subjects' eyes, it was found that SS-OCT PCT, Pentacam gray value, low vision quality of life questionnaire (LVQ questionnaire) for distance vision, and mobility and lighting dimension were significantly correlated with the PCO score (P = 0.012, P = 0.001, P = 0.005, respectively).ConclusionThe region segmentation and adaptive threshold algorithm of posterior capsule image will accurately quantify the posterior capsule. Computer aided quantifications of posterior capsule are of great significance in the early surgical decision-making of PCO. The average occurrence time of most PCO was around 34 months, and the severity of PCO worsened with increasing postoperative time.

Hybrid iteration and optimization-based three-dimensional reconstruction for space non-cooperative targets with monocular vision and sparse lidar fusion

The measurement of reference pose and geometric size for space non-cooperative targets is an essential premise of on-orbit servicing. In this paper, a three-dimensional (3-D) reconstruction and autonomous geometric parameter identification method is proposed. An autonomous region segmentation algorithm with monocular vision and sparse lidar fusion is introduced, which realizes the multi-source match of rich features in images and scale information in point clouds. In order to densify the sparse point cloud, a multi-view fusion model is constructed with multi-objective optimization, which realizes the complementation of 3-D information. Further, a 3-D reconstruction method with hybrid iteration and optimization is proposed. Unlike the method with random sampling, the proposed method can obtain the optimal solution while reducing the influence of outliers. In other to verify the effectiveness of the proposed method, a complete set of simulation and experimental systems with modular and scalable software is built. The results show that the accuracy and stability are improved by no less than 27.88% and 38.99%, respectively. Besides, the inference speed is enhanced by at least one time. The proposed method can accurately and quickly identify the geometric parameters of targets with a cubical body, which will be used for motion estimation. Moreover, it can be extended to targets with a polyhedral body.

Path planning and pose correction of robot laser cleaning process for specific surfaces of parts

The purpose of this paper is to solve a series of problems, such as the difficulty in accurately controlling the specific surfaces, and the interference between the laser beam and the part, when the laser is used to clean mechanical parts with complex surfaces by using robots. Firstly, a region segmentation algorithm of the stereo lithography (STL) model was designed, and the feature surfaces of the model were obtained by using the dihedral angle as the criterion. Then the path points which could easily cause interference between the laser beam and the part to be cleaned were detected by the method of ray intersection, and the pose of the robot was modified according to the normal vectors of two non-interference points that were adjacent to the continuous interference points respectively. Finally, the proposed method was validated by scanning path visualization and robot laser cleaning process experiments. The generated path in the unit of feature surface according to the variation components of normal vectors proved the effectiveness of the STL model region segmentation algorithm based on dihedral angle division criteria; the change of the normal vectors of the path points on the current feature surface was continuous and smooth, which meant that the path points did not cause sudden change of the robot pose after the pose correction. The robot cleaned the specific surfaces of the test workpiece safely and effectively in the actual laser cleaning experiments, and the effective cleaning area reached 96.87%, 99.56%, 94.53% and 91.62% of the target cleaning area, which further proved the effectiveness of this study. The method is also useful for the path planning of robots in other processing fields.

Comparison of Object Region Segmentation Algorithms of PCB Defect Detection

Comparison of Object Region Segmentation Algorithms of PCB Defect Detection

Target region extraction and segmentation algorithm for reflective tomography Lidar image

Target region extraction and segmentation algorithm for reflective tomography Lidar image

Multi-surface defect detection for universal joint bearings via multimodal feature and deep transfer learning

ABSTRACT Surface defect detection by machine vision has received increased attention concerning the quality control of universal joint bearings (UJB). The defect distribution and counting information are important for product quality optimisation. However, vision defect detection for UJB remains a challenging task due to the diversity of background textures and defect characters on multiple surfaces. In this study, a multi-surface defect detection (MSDD) method consisting of region segmentation, feature extraction and detection is proposed. First, defect regions are accurately localised by the proposed adaptive defect region segmentation algorithm, which suppresses the inference of background variety. Then, a novel defect feature named multimodal fusion shape descriptor that integrates the global information and local information of defects is constructed to generate the discriminative defect representation. Finally, a defect feature extraction ability transfer strategy based on the transfer learning mechanism is proposed to address the problem of insufficient defect samples. The experimental results show that our method achieves the best accuracy of 94.8% macro-F1 and processes 28 defect images per second. Besides, the application effect in the practical production line indicates that our method meets the accuracy and real-time requirements of MSDD for UJB.

Diabetic foot ulcer segmentation using logistic regression, DBSCAN clustering and mathematical morphology operators

Digital images are used for evaluation and diagnosis of a diabetic foot ulcer. Selecting the wound region (segmentation) in an image is a preliminary step for subsequent analysis. Most of the time, manual segmentation isn't very reliable because specialists could have different opinions over the ulcer border. This fact encourages researchers to find and test different automatic segmentation techniques. This paper presents a computer-aided ulcer region segmentation algorithm for diabetic foot images. The proposed algorithm has two stages: ulcer region segmentation, and post-processing of segmentation results. For the first stage, a trained machine learning model was selected to classify pixels inside the ulcer's region, after a comparison of five learning models. Exhaustive experiments have been performed with our own annotated dataset from images of Cuban patients. The second stage is needed because of the presence of some misclassified pixels. To solve this, we applied the DBSCAN clustering algorithm, together with dilation, and closing morphological operators. The best-trained model after the post-processing stage was the logistic regressor (Jaccard Index $0.81$, accuracy $0.94$, recall $0.86$, precision $0.91$, and F1 score $0.88$). The trained model was sensitive to irrelevant objects in the scene, but the patient foot. Physicians found these results promising to measure the lesion area and to follow-up the ulcer healing process over treatments, reducing errors.

Adoption and Safety Evaluation of Comfortable Nursing by Mobile Internet of Things in Pediatric Outpatient Sedation.

The aim of study was to explore the application effect and safety of comfortable nursing based on optimized mobile Internet of Things (MIoT) in the clinical sedation and diagnosis of mycoplasma pneumoniae pneumonia (MPP) in children. A total of 70 children with MPP admitted to the respiratory clinic of hospital were randomly selected and divided into a control group (comfortable nursing mode) and an observation group (comfortable nursing mode based on optimized MIoT), with 35 cases in each. The nursing effects and safety were compared between groups. The results showed that the node jitter rate, delivery success rate, and congestion times of the multilayered sensing algorithm were better than those of the mobile relay area segmentation algorithm and the wedge merge-energy hole elimination area segmentation algorithm. The CD-RISC resilience score of the observation group ((94.72 ± 1.58) points), the proportion of children with Frankl-3 and 4 points (90%), and the comfort level ((95.01 ± 5.68) points) were higher than those of the control group ((64.12 ± 1.62) points, (33.33%), and (55.23 ± 6.18) points) (P < 0.05). After treatment, the proportion of children with HRCT image lesions in the observation group was lower than that in the control group (P < 0.05). After treatment, the FEV1 ((85.71 ± 5.23) % vs. (68.26 ± 5.90) %) and FEV1/FVC ((74.22 ± 2.12) % vs (64.38 ± 2.34) %) of the observation group were significantly better than those of the control group (P < 0.05). The results showed that the incidence of adverse reactions in the observation group (14%) was significantly lower than that in the control group (46%) (P < 0.05). MIoT-assisted comfort nursing based on multilayer perception region segmentation algorithm can more effectively relieve the emotions of children in MPP outpatient department during sedation and diagnosis and treatment, improve the therapeutic effect and safety, and is worthy of clinical promotion.

Concrete Application of Computer Virtual Image Technology in Modern Sports Training.

The efficiency of basketball shooting training has always been an urgent problem to be solved in basketball sports. Applying computer virtual image technology to basketball sports training is the key to improve basketball shooting ability and improve the effect of basketball sports training. Therefore, based on the design of basketball shooting automatic recognition system based on the background difference method, this study puts forward the specific application of computer virtual image technology in modern sports training. By analyzing the techniques of image denoising, image detection, and image calibration, a goal detection algorithm for modern sports basketball shooting training is designed. Firstly, the camera is used for image capture, the RGB image is converted into gray image, and the median filter is used to suppress the noise in the image. Then, the background difference method is used to detect the moving region, and the background modeling is combined with the mean method. After obtaining the background reference model, the image is differentiated, the gray image after image difference is binarized, and then the binary image is postprocessed by morphological middle closure operation. Finally, the image calibration technology is used to extract the basketball feature information. Through the region segmentation algorithm, the basketball shooting part is segmented and judged, so as to realize the basketball shooting training goal detection. The experimental results show that the proposed method has a good effect on basketball shooting training goal detection and can effectively improve the detection accuracy and efficiency.

Computed tomography of ground glass nodule image based on fuzzy C-means clustering algorithm to predict invasion of pulmonary adenocarcinoma

Early detection of lung cancer in computed tomography (CT) can significantly improve the survival rate of patients. A part of lung cancer early appeared in the form of ground glass nodule (GGN), and its early detection requires the help of a computer-aided algorithm. This study aimed to explore the CT features of GGN based on a fuzzy C-means (FCM) clustering algorithm in predicting the invasion of pulmonary adenocarcinoma. In this study, the lung parenchyma from 65 patients with GGN was segmented based improved FCM cluster algorithm. The region segmentation algorithm removed the useless area, and critical features of GGN extracted the suspicious area. After using the improved algorithm, the effect of region segmentation was superior; the simulation results show that the algorithm can remove blood vessels' “line” or branching structure well; the improved FCM had good real-time performance and robustness to noise. From the segmented GGN, the invasive adenocarcinoma (IA) had a larger size and higher CT attenuation than that of non-IA lesions. The results show that the accurate analysis of CT features of GGN based on the FCM clustering algorithm could be helpful in distinguish non-IA from IA, which is beneficial to determine strategies of follow-up or appropriate management for GGN.

Application of Artificial Intelligence in Cardiovascular Imaging.

During the last two decades, as computer technology has matured and business scenarios have diversified, the scale of application of computer systems in various industries has continued to expand, resulting in a huge increase in industry data. As for the medical industry, huge unstructured data has been accumulated, so exploring how to use medical image data more effectively to efficiently complete diagnosis has an important practical impact. For a long time, China has been striving to promote the process of medical informatization, and the combination of big data and artificial intelligence and other advanced technologies in the medical field has become a hot industry and a new development trend. This paper focuses on cardiovascular diseases and uses relevant deep learning methods to realize automatic analysis and diagnosis of medical images and verify the feasibility of AI-assisted medical treatment. We have tried to achieve a complete diagnosis of cardiovascular medical imaging and localize the vulnerable lesion area. (1) We tested the classical object based on a convolutional neural network and experiment, explored the region segmentation algorithm, and showed its application scenarios in the field of medical imaging. (2) According to the data and task characteristics, we built a network model containing classification nodes and regression nodes. After the multitask joint drill, the effect of diagnosis and detection was also enhanced. In this paper, a weighted loss function mechanism is used to improve the imbalance of data between classes in medical image analysis, and the effect of the model is enhanced. (3) In the actual medical process, many medical images have the label information of high-level categories but lack the label information of low-level lesions. The proposed system exposes the possibility of lesion localization under weakly supervised conditions by taking cardiovascular imaging data to resolve these issues. Experimental results have verified that the proposed deep learning-enabled model has the capacity to resolve the aforementioned issues with minimum possible changes in the underlined infrastructure.

A full‐reference stereoscopic image quality assessment index based on stable aggregation of monocular and binocular visual features

In stereoscopic image quality assessment, human visual system has been universally taken into account to detect perceptual characteristics. A novel full-reference stereoscopic image assessment metric by considering both monocular and binocular visual features of human visual system is proposed. In particular, a new region segmentation algorithm is firstly proposed to divide 3D images into occluded and non-occluded regions. The just noticeable difference model is employed on the occluded regions to formulate the monocular vision, while the binocular just noticeable difference model is applied to the non-occluded regions to reveal the binocular vision of the human visual system. In the proposed region segmentation, disparity information and Euclidean distance between stereo pairs are both adopted to solve the unstable segmentation problem of traditional methods. A new pooling strategy based on global edge features is then presented to aggregate the just noticeable difference and binocular just noticeable difference evaluation maps. In addition, some local image features as supplementary of just noticeable difference to describe visual characteristics of the human visual system are also extracted. Finally, an overall quality score is calculated based on the above-mentioned features to measure the visual quality of distorted stereo pairs. Experimental results show that the proposed metric achieves high consistency with the human visual system, and outperforms state-of-the-art algorithms on stereoscopic image quality assessment.

工程制图习题答题区域分割算法研究

工程制图习题答题区域分割算法研究