Region Of Interest Extraction Research Articles

Anomaly Detection in Banking Using A Video Surveillance System

An anomaly detection scheme is proposed for encrypted video bitstream with secure video encryption. Human beings are recognized by their unique facial characteristics. In the present work time based movement and face recognition approach will be implement to detect person in unwanted time.In video sharing , ROI (Region of Interest) extraction can be implement to detect the region to hide. An efficient encryption technique is used to encrypt the extracted region.

High-resolution SAR images segmentation using NSCT denoising and QIGA based parameters selection of PCNN model

The high-resolution synthetic aperture radar (SAR) images usually contain inhomogeneous coherent speckle noises. For the high-resolution SAR image segmentation with such noises, the conventional methods based on pulse coupled neural networks (PCNN) have to face heavy parameters with a low efficiency. In order to solve the problems, this paper proposes a novel SAR image segmentation algorithm based on non-subsampling Contourlet transform (NSCT) denoising and quantum immune genetic algorithm (QIGA) improved PCNN models. The proposed method first denoising the SAR images for a pre-processing based on NSCT. Then, by using the QIGA to select parameters for the PCNN models, such models self-adaptively select the suitable parameters for segmentation of SAR images with different scenes. This method decreases the number of parameters in the PCNN models and improves the efficiency of PCNN models. At last, by using the optimal threshold to binary the segmented SAR images, the small objects and large scales from the original SAR images will be segmented. To validate the feasibility and effectiveness of the proposed algorithm, four different comparable experiments are applied to validate the proposed algorithm. Experimental results have shown that NSCT pre-processing has a better performance for coherent speckle noises suppression, and QIGA-PCNN model based on denoised SAR images has an obvious segmentation performance improvement on region consistency and region contrast than state-of-the-arts methods. Besides, the segmentation efficiency is also improved than conventional PCNN model, and the level of time complexity meets the state-of-the-arts methods. Our proposed NSCT+QIGA-PCNN model can be used for small object segmentation and large scale segmentation in high-resolution SAR images. The segmented results will be further used for object classification and recognition, regions of interest extraction, and moving object detection and tracking.

Cancellable face template algorithm based on speeded-up robust features and winner-takes-all

Features such as face, fingerprint, and iris imprints have been used for authentication in biometric system. The toughest feature amongst these is the face. Extracting a region with the most potential face features from an image for biometric identification followed by illumination enhancement is a commonly used method. However, the region of interest extraction followed by illumination enhancement is sensitive to image face feature displacement, skewed image, and bad illumination. This research presents a cancell able face image algorithm built upon the speeded-up robust features method to extract and select features. A speeded-up robust feature approach is utilised for the image’s features extraction, while Winner-Takes-All hashing is utilised for match-seeking. Finally, the features vectors are projected by utilising a random form of binary orthogonal matrice. Experiments were conducted on Yale and ORL datasets which provide gray scale images of sizes 168 × 192 and 112 × 92 pixels, respectively. The execution of the proposed algorithm was measured against several algorithms using equal error rate metric. It is found that the proposed algorithm produced an acceptable performance which indicates that this algorithm can be used in biometric security applications.

Lumbar disk 3D modeling from limited number of MRI axial slices

This paper studies the problem of clinical MRI analysis in the field of lumbar intervertebral disk herniation diagnosis. It discusses the possibility of assisting radiologists in reading the patients MRI images by constructing a 3D model for the region of interest using simple computer vision methods. We use axial MRI slices of the lumbar area. The proposed framework works with a very small number of MRI slices and goes through three main stages. Namely, the region of interest extraction and enhancement, inter-slice interpolation, and 3D model construction. We use the Marching Cubes algorithm to construct the 3D model of the the region of interest. The validation of our 3D models is based on a radiologist’s analysis of the models. We tested the proposed 3D model construction on 83 cases and We have a 95% accuracy according to the radiologist evaluation. This study shows that 3D model construction can greatly ease the task of the radiologist which enhances the working experience. This leads eventually to more accurate and easy diagnosis process.

Breast tumor detection and classification based on density

Breast cancer is the most widely disease in women and is considered one of the biggest causes of death in women. Early detection, classification, and diagnosis of this cancer are essential to reduce the death cases. The efficiency of the diagnosis of breast cancer depends on the accuracy of the segmentation of the tumor and its classification. Before the segmentation of mammogram images, it is necessary to extract the Region of Interest (ROI) because it can assume a significant job in the efficiency of tumor detection. In this study, the method of automatic extraction of ROI from the mammogram image is presented. After that, the preprocessing of both mammogram and MR images has been done which removes the noise and enhances the contrast of the images. On the processed images, segmentation of tumors from mammograms as well as Magnetic Resonance Imaging (MRI) breast images has been performed using automatic seed point extraction and threshold calculation in Seeded Region Growing (SRG). After the successful detection of breast tumor area, classification of the tumor as benign or malignant has been done. The test was performed on three publicly available data sets; RIDER breast MR Images, Mammographic Image Analysis Society (MIAS), and Digital Database for Screening Mammography (DDSM). The test showed that classification accuracy is better than previously accessible cutting edge techniques, which is 91.4%. The proposed method not only detects the shape and size of the tumor from both mammograms and MR Images but also efficiently classifies the tumor as benign or malignant.

Task-Driven Regional Saliency Analysis Based on a Global–Local Feature Assembly Network in Complex Optical Remote Sensing Scenes

Saliency analysis is an essential task in computer vision and aims to generate distinguishing foreground features from background features. However, due to complex structure distributions in large-scale optical remote sensing scenes, generating effective feature descriptions for regional saliency analysis is challenging. Therefore, in this study, we proposed a novel global-local feature assembly method called GLFA-Net based on a convolution neural network (CNN) that can adaptively learn an effective feature representation for regional saliency analysis to achieve the region-of-interest (ROI) (e.g., aircraft carrier, airport, and urban area) extraction from complex optical remote sensing images. In addition, we also collected these complex optical remote sensing scene images from Google Earth and DOTA data sets to demonstrate the effectiveness of the proposed method. Finally, experimentation shows that the proposed regional saliency analysis method can produce better ROI extraction performance than other methods, reaching a 0.057 mean absolute error (MAE), a 0.703 Kappa coefficient, and a 0.735 F1-score.

OP0059 AUTOSCORA: DEEP LEARNING TO AUTOMATE SCORING OF RADIOGRAPHIC PROGRESSION IN RHEUMATOID ARTHRITIS

Background:The prevention of joint destruction is an important goal in the management of rheumatoid arthritis (RA) and a key endpoint in drug trials. To quantify structural damage in radiographs, standardized scoring systems1, such as the Sharp/van der Heijde (SvdH) score2, which separately assesses joint space narrowing (JSN) and erosions, have been developed. However, application of these scores is time-consuming, requires specially trained staff, and results are subject to considerable intra- and inter-reader variability1. This makes their application poorly feasible in clinical practice and limits their reliability in clinical trials.Objectives:We aim to develop a fully automated deep learning-based scoring system of radiographic progression in RA to facilitate the introduction of quantitative joint damage assessment into daily clinical practice and circumvent inter-reader variability in clinical trials.Methods:5191 hand radiographs and their corresponding SvdH JSN scores from 640 adult patients with RA without visible joint surgery were extracted from the picture archive of a large tertiary hospital. The dataset was split, on a patient level, into training (2207 images/270 patients), validation (1150/133), and test (1834/237) sets. Joints were automatically localized using a particular deep learning model3which utilizes the local appearance of joints combined with information on the spatial relationship between joints. Small regions of interest (ROI) were automatically extracted around each joint. Finally, different deep learning architectures were trained on the extracted ROIs using the manually assigned SvdH JSN scores as ground truth (Fig. 1). The best models were chosen based on their performance on the validation set. Their ability to assign the correct SvdH JSN scores to ROIs was assessed using the unseen data of the test set.Fig. 1.3-step approach to automated scoring: joint localization, ROI extraction, JSN scoring.Results:ROI extraction was successful in 96% of joints, meaning that all structures were visible and joints were not malrotated by more than 30 degrees. For JSN scoring, modifications of the VGG164architecture seemed to outperform adaptations of DenseNet5. The mean obtained accuracy (i.e., the percentage of joints to which the human reader and our system assigned the same score) for MCP joints was 80.5 %, that for PIP joints was 72.3 %. In only 1.8 % (MCPs) and 1.7 % (PIPs) of cases did the predicted score differ by more than one point from the ground truth (Fig. 2).Fig. 2.Confusion matrices of automatically assigned scores (‘predicted score’) vs. the human reader ground truth (‘true score’).Conclusion:Although a number of previous efforts have been published, none have succeeded in replacing manual scoring systems at scale. To our knowledge, this is the first work that utilizes a dataset of adequate size to apply deep learning to automate JSN scoring. Our results are, even in this early version, in good agreement with human reader ground truth scores. In future versions, this system can be expanded to the detection of erosions and to all joints contained in the SvdH score.

Breast Infrared Thermography Segmentation Based on Adaptive Tuning of a Fully Convolutional Network

Accurate segmentation of Breast Infrared Thermography is an important step for early detection of breast pathological changes. Automatic segmentation of Breast Infrared Thermography is a very challenging task, as it is difficult to find an accurate breast contour and extract regions of interest from it. Although several semi-automatic methods have been proposed for segmentation, their performance often depends on hand-crafted image features, as well as preprocessing operations. In this work, an approach to automatic semantic segmentation of the Breast Infrared Thermography is proposed based on end-to-end fully convolutional neural networks and without any pre or post-processing. The lack of labeled Breast Infrared Thermography data limits the complete utilization of fully convolutional neural networks. The proposed model overcomes this challenge by applying data augmentation and two-tier transfer learning from bigger datasets combined with adaptive multi-tier fine-tuning before training the fully convolutional neural networks model. Experimental results show that the proposed approach achieves better segmentation results: 97.986% accuracy; 98.36% sensitivity and 97.61% specificity compared to hand-crafted segmentation methods. This work provided an end-to-end automatic semantic segmentation of Breast Infrared Thermography combined with fully convolutional networks, adaptive multi-tier fine-tuning and transfer learning. Also, this work was able to deal with challenges in applying convolutional neural networks on such data and achieving the state-of-the-art accuracy.

Fingerprint liveness detection based on guided filtering and hybrid image analysis

Fingerprints are widely used for biometric recognition. However, many spoofing attacks based on an artificially made fingerprint occur. In this study, the authors propose an approach to detect fingerprint liveness which uses the guided filtering and hybrid image analysis. This study deals with the problem of ignoring the contribution that is brought by the sharp features when analysing the denoised image. The method described utilises both the enhanced sharp features and denoised features from the hybrid images to get better results. The input fingerprint is pre-processed by region of interest extraction and then is filtered by a guidance image for obtaining the denoised image. Then, histogram equalisation is introduced to eliminate the impact of illumination condition. The authors extract the co-occurrence of adjacent local binary pattern features from both the cropped images and the denoised images. Whilst concatenating both the features together to form a long feature, t-Distributed Stochastic Neighbour Embedding is applied to reduce the data dimension. The authors consider the fingerprint liveness detection as a two-class classification problem and use support vector machine with radial basis function kernel to solve this problem. The authors evaluate the experiments on three benchmark data sets. Experimental results demonstrate that the accuracy of the proposed method can outperform most of the state-of-art methods.

An efficient method for extraction and recognition of bangla characters from vehicle license plates

Recognition of characters from vehicle license plate plays a vital role in vehicle tracking, controlling, and maintaining traffic on the roads and high ways. This paper presents an automatic system to detect plate area and recognize characters. The research is carried out on Bangladeshi vehicle license plates because this is a demandable research area in the present age. The method has been implemented using the five consecutive steps such as license plate detection, extraction, character localization, extraction of characters, and recognition of extracted characters. We have introduced an algorithm to detect the Region of Interest (ROI) from the input image. It is performed by applying morphological operation and histogram analysis of vertical and horizontal projection profiles. A dynamic threshold is used to filter out both horizontal and vertical histogram values and ROI is extracted from the license plate image using different geometric properties such as area, bounding box, and aspect ratio. Character segmentation is performed by applying the method of connected component analysis and bounding box technology. Finally, character recognition is carried out using Support Vector Machine (SVM) classifier where extracted Histogram of Oriented Gradient (HOG) features are used as input. The proposed algorithm is applied to 630 images of license plates of different categories of vehicles and achieved 91% accuracy in the extraction of ROI and 94.6% accuracy in the extraction of characters. We have also achieved 100% accuracy in recognition of Bangla characters from a total of 4725 Bangla characters extracted from 630 images of license plates.

Small target detection based on automatic ROI extraction and local directional gray&entropy contrast map

As a special problem in object recognition, how to detect the small target from complex background fast and precisely is kept an open topic. An excellent algorithm should have ideal Pd (Probability of detection) as well as lower Fa (False alarm) provided that no missing detection. Taking it as the goal, in this paper, we put forward an enhanced version based on LDG&ECM (Local Directional Gray&Entropy Contrast Map). To enhance the discrimination of the target with the background clutter, the saliency feature of “contrast” is remearsured by associating the “gray intensity” with “entropy”. Also, with the introduction of “directivity” in the calculation, the stubborn clutter edge with similar property to the target is removed effectively. Moreover, to speed up the algorithm, the cGANs (Conditional Generative Adversarial Networks) is first introduced to extract ROI (Region of Interest) automatically. Based on DNNs, the traditional extraction methods relying on empirical threshold is improved to an end-to-end way. Additional, to detect the multiple targets in heterogeneous background, the original image is segmented into multiple homogeneous parts by OTSU. And then, a novel local adaptive threshold decision making is designed. By this way, all of the real targets with prominent local extreme property are detected and the false alarm caused by local extreme is avoided due to the balance of global information. It is verified that the proposed automation and adaptively approach has significant performance improvement compared with the state-of-the-art algorithms in various scenarios.

ROI Extraction Based on Multiview Learning and Attention Mechanism for Unbalanced Remote Sensing Data Set

With an increasing number of remote sensing images (RSIs), the automatic region of interest (ROI) extraction based on convolutional neural networks (CNNs) has attracted much interest in recent years. Although fully supervised CNN-based methods have shown superiority in the field of object extraction, pixelwise annotations are expensive and time-consuming. Moreover, due to the unstable distribution of ROIs in complex landscapes, the ratios of foreground and background areas are quite different in RSIs. Training CNNs with such unbalanced data sets lead to over-fitting and low accuracy. In this article, we propose a framework that combines multiview learning and attention mechanism (MLAM) to solve the above mentioned problems. First, we develop a CNN-based weakly supervised method with a weight-balanced loss function to solve the problems caused by an unbalanced data set. It also helps to generate imagewise saliency maps by computing the gradient maps with respect to the input images. Then, we design a multiview strategy to dramatically reduce the missing inspection. Finally, we design a feedback attention mechanism based on the stage neighbor binary pattern to further modify the extraction result. In summary, the proposed framework achieves pixelwise ROI extraction under imagewise annotations through data-driven ML and a knowledge-driven visual attention mechanism. We evaluate the performance of the MLAM framework on two challenging data sets with complex backgrounds. The experimental results indicate that the proposed framework can achieve better performance than other eight ROI extraction models for unbalanced remote sensing data sets.

A review of mammographic region of interest classification

AbstractEarly detection of breast cancer is important and highly valuable in clinical practice. X‐ray mammography is broadly used for prescreening the breast and is also attractive due to its noninvasive nature. However, experts can misdiagnose a significant proportion of the cases, which may either cause redundant examinations or cancer. In order to reduce false positive and negative rates of mammography screening, computer‐aided breast cancer detection has been studied for more than 30 years and many methods have been proposed by the researchers. In this review, region of interest (ROI) classification methods, which operate on a predefined or segmented ROIs with a focus on mass classification are surveyed. A total of 72 high quality journal and conference papers are selected from the Web of Science (WOS) database that meet several inclusion criteria. A comparative analysis is provided based on ROI extraction methods, data sets and machine learning techniques employed, the prediction accuracies, and usage frequency statistics. Based on the performances obtained on publicly available data sets, the ROI classification problem from mammogram images can be considered as approaching to be solved. Nonetheless, it can still be used as complementary information in breast cancer detection from the whole mammograms, which has room for improvement.This article is categorized under: Application Areas > Science and Technology Technologies > Machine Learning Technologies > Classification

FPGA-Based ROI Encoding for HEVC Video Bitrate Reduction

The explosive growth of video applications has produced great challenges for data storage and transmission. In this paper, we propose a new ROI (region of interest) encoding solution to accelerate the processing and reduce the bitrate based on the latest video compression standard H.265/HEVC (High-Efficiency Video Coding). The traditional ROI extraction mapping algorithm uses pixel-based Gaussian background modeling (GBM), which requires a large number of complex floating-point calculations. Instead, we propose a block-based GBM to set up the background, which is in accord with the block division of HEVC. Then, we use the SAD (sum of absolute difference) rule to separate the foreground block from the background block, and these blocks are mapped into the coding tree unit (CTU) of HEVC. Moreover, the quantization parameter (QP) is adjusted according to the distortion rate automatically. The experimental results show that the processing speed on FPGA has reached a real-time level of 22 FPS (frames per second) for full high-definition videos ([Formula: see text]), and the bitrate is reduced by 10% on average with stable video quality.

Research on Lung Nodule Recognition Based on Stacking Algorithm

In recent years, the incidence of cancer has been increasing year by year, and lung cancer has ranked the first place. In order to detect lung nodules quickly and accurately, since the lung CT images in the data set used in this experiment are original medical images, image preprocessing is first required, including noise reduction, segmentation and region of interest extraction, Therefore, this paper proposes a lung nodule recognition method based on Stacking algorithm to detect lung nodules quickly and accurately. Secondly, the three-dimensional gray co-occurrence matrix texture features of the region of interest are extracted. Finally, the Stacking evaluation model is used to identify and classify lung nodules.The experimental results show: Under ten-fold cross-validation, the training accuracy rate and the test accuracy rate can reach 90.9%,82.3%, respectively.This method is feasible for pulmonary nodule recognition and provides data reference value for doctors' auxiliary diagnosis.

Region of interest extraction for image quality assessment

Print quality (PQ) is most important in the printing industry. To detect and analyze print defects is an effective solution to improve print quality. As the different types of print defects appear in different regions of interest (ROI) in the digital image of a scanned page, extracting the different ROIs helps to detect and analyze the printer defect. This paper proposes a method to extract different ROIs based on the digital image object map [1], which includes three different labels: raster (images or pictures), vector (background and smooth gradient color areas), and symbol (symbols and texts). Our ROI extraction method will extract four kinds of ROIs based on these three labeled objects. So we need to distinguish the background area and smooth gradient color area (color vector) from other vector objects. The process of the ROI extraction method includes four parts; and each part will extract one kind of ROI. For the color vector and background ROI extraction part, we develop two approaches: one is to obtain the maximum area rectangular ROI; and the other approach is to extract the deepest rectangular ROI. With both of these two methods, we use a greedy algorithm to gather additional useful ROIs. In the final result of the ROI extraction process, we only save the left top and right bottom positions for each ROI. In the end, we design a Matlab GUI Tool and label the ROI ground truth manually. We calculate the intersection over union (IoU)) between the ROI extraction result and the ROI manually labeled ground truth to evaluate our ROI extraction algorithm, and check whether it is good enough to crop different ROIs from the image of the scanned page to detect and analyze print defects.

Parallel extraction of Regions‐of‐Interest from social media data

SummaryGeotagged data gathered from social media can be used to discover places‐of‐interest (PoIs) that have attracted many visitors. Since a PoI is generally identified by geographical coordinates of a single point, it is hard to match it with people trajectories. Therefore, we define an area, called region‐of‐interest (RoI), represented by the boundaries of a PoI. The main goal of this study is to discover RoIs from PoIs using spatial data mining techniques. In this paper, we propose a new parallel method for extracting RoIs from social media datasets. It consists of two main steps: (i) automatic keyword extraction and data grouping and (ii) parallel RoI extraction. The first step extracts keywords identifying the PoIs; these keywords are used to group social media items according to the places they refer to. The second step uses a Parallel Clustering Approach (ParCA) of spatial dataset to identify RoIs. ParCA exploits a parallel execution of DBSCAN on subsets of data to generate subclusters on each processing node and then merge overlapping subclusters to form global clusters. ParCA was implemented using the MapReduce model. Experiments performed over a set of PoIs in the city of Rome using social media data show that our approach is highly scalable and reaches an accuracy of 79% in detecting RoIs. On a parallel computer with 50 cores, we obtained a speedup of 52 by processing large datasets divided into 32 splits, compared with the execution time registered when each dataset is not partitioned.

New Algorithm for Palm Print Detection

Advances in technology are increasing the demand for the biometric technology of palm print detection. This work presents a region of interest segmentation algorithm for palm prints based on Hue Saturation Brightness. The aim of this work was to detect palm print region of interest based on skin color and geometric features. After a color palm image is read as a JPEG file, the algorithm passes through three main stages of preprocessing: palm localization, region of interest localization and region of interest extraction. The preprocessing stage consists of three steps: conversion to Hue Saturation Brightness, skin color modeling, and de-noising. The second main stage of palm localization consists of clipping from the left, from the right, from the top and from bottom. The third stage consists of the two steps of determining and extracting region of interest. The algorithm was tested on a dataset of 180 palm images from the Institute of Automation Chinese Academy of Sciences dataset, which contains an equal number of right-hand and left-hand palm images. The results showed that the Hue Saturation Brightness color space for skin detection was superior to other recognition methods, with an accuracy rate of 98.8%.

Dynamic Downscaling Segmentation for Noisy, Low-Contrast in Situ Underwater Plankton Images

Finding and segmenting objects in noisy low-contrast in situ underwater plankton images is challenging because of the difficulty of separating potential plankton objects from the complex background and numerous and diverse other particles. In the present study, a dynamic downscaling model was developed to rapidly extract complete and clean regions of interest (ROIs) from images with highly variable content and quality. The original image was downscaled, and dynamic segmentation was performed in a scale pyramid space to ensure the integrity of weak targets based on local two-dimensional (2D) entropy parameters. Subsequently, a series of local thresholds and clustering gradients was examined iteratively for ROI selection. The performance of the local 2D entropy parameters relative to water turbidity (in a scattering medium and with high background noise) and image size was examined. To suppress the background and increase the sharpness of potential targets, a sharpness descriptor and gradient clustering were employed. The method was compared with the currently commonly used local threshold-based Sauvola segmentation using the same set of images. The results showed that the proposed method improves the ROI extraction accuracy and reduces oversegmentation for in situ underwater plankton images. It was concluded that the proposed method is a fast and robust segmentation technique and could facilitate the deployment of in situ plankton imaging systems for process-based research and routine plankton monitoring.

Region-of-Interest Based Transfer Learning Assisted Framework for Skin Cancer Detection

Melanoma is considered the most serious type of skin cancer. All over the world, the mortality rate is much high for melanoma in contrast with other cancer. There are various computer-aided solutions proposed to correctly identify melanoma cancer. However, the difficult visual appearance of the nevus makes it very difficult to design a reliable Computer-Aided Diagnosis (CAD) system for accurate melanoma detection. Existing systems either uses traditional machine learning models and focus on handpicked suitable features or uses deep learning-based methods that use complete images for feature learning. The automatic and most discriminative feature extraction for skin cancer remains an important research problem that can further be used to better deep learning training. Furthermore, the availability of the limited available images also creates a problem for deep learning models. From this line of research, we propose an intelligent Region of Interest (ROI) based system to identify and discriminate melanoma with nevus cancer by using the transfer learning approach. An improved k-mean algorithm is used to extract ROIs from the images. These ROI based approach helps to identify discriminative features as the images containing only melanoma cells are used to train system. We further use a Convolutional Neural Network (CNN) based transfer learning model with data augmentation for ROI images of DermIS and DermQuest datasets. The proposed system gives 97.9% and 97.4% accuracy for DermIS and DermQuest respectively. The proposed ROI based transfer learning approach outperforms existing methods that use complete images for classification.