Traitement du Signal

Journal Information
ISSN / EISSN : 0765-0019 / 1958-5608
Total articles ≅ 462
Current Coverage
SCOPUS
SCIE
COMPENDEX
Filter:

Latest articles in this journal

Vasileios Papageorgiou
Traitement du Signal, Volume 38, pp 547-554; https://doi.org/10.18280/ts.380302

Abstract:
Brain tumor detection or brain tumor classification is one of the most challenging problems in modern medicine, where patients suffering from benign or malignant brain tumors are usually characterized by low life expectancy making the necessity of a punctual and accurate diagnosis mandatory. However, even today, this kind of diagnosis is based on manual classification of magnetic resonance imaging (MRI), culminating in inaccurate conclusions especially when they derive from inexperienced doctors. Hence, trusted, automatic classification schemes are essential for the reduction of humans’ death rate due to this major chronic disease. In this article, we propose an automatic classification tool, using a computationally economic convolutional neural network (CNN), for the purposes of a binary problem concerning MRI images depicting the existence or the absence of brain tumors. The proposed model is based on a dataset containing real MRI images of both classes with nearly perfect validation-testing accuracy and low computational complexity, resulting a very fast and reliable training-validation process. During our analysis we compare the diagnostic capacity of three alternative loss functions, validating the appropriateness of cross entropy function, while underlining the capability of an alternative loss function named Jensen-Shannon divergence since our model accomplished nearly excellent testing accuracy, as with cross-entropy. The multiple validation tests applied, enhancing the robustness of the produced results, render this low-complexity CNN structure as an ideal and trustworthy medical aid for the classification of small datasets.
Yuanguo Liu,
Traitement du Signal, Volume 38, pp 599-605; https://doi.org/10.18280/ts.380307

Abstract:
The effect of motion posture recognition hinges on the accurate description of motion postures with effective feature information. This study introduces Wronskian function to improve the denoising ability of visual background extractor (ViBe) algorithm, and thus acquires relatively clear motion targets. Then, a multi-feature fusion motion posture feature model was developed based on genetic algorithm (GA). Specifically, GA was called to optimize and fuse the extracted feature information, while a fitness function was constructed based on the mean variance ratio, and used to select the feature information with high inter-class discriminability. Taking support vector machine (SVM) as the classifier, a multi-class classifier was designed by one-to-one method for the classification and recognition of motion postures. Through experiments, our model was proved highly accurate in motion posture recognition.
, Ganjikunta Ganesh Kumar, Pandya Vyomal Naishadkumar, Sarvade Pedda Subba Rao
Traitement du Signal, Volume 38, pp 731-738; https://doi.org/10.18280/ts.380320

Abstract:
Diagnosing chronic obstructive pulmonary disease (COPD) from lung sounds is time consuming, onerous, and subjective to the expertise of pulmonologists. The preliminary diagnosis of COPD is often based on adventitious lung sounds (ALS). This paper proposes to objectively analyze the lung sound signals associated with COPD. Specifically, empirical mode decomposition (EMD), a data adaptive signal decomposition technique suitable for analyzing non-stationary signals, was adopted to decompose non-stationary lung sound signals. The use of EMD on lung sound signal results in intrinsic mode functions (IMFs), which are symmetric and band limited. The analytic IMFs were then computed through the Hilbert transform, which reveals the instantaneous frequency content of each IMF. The Hilbert transformed signal is analytic, and has a complex representation containing real and imaginary parts. Next, the central tendency measure (CTM) was introduced to quantify the circular shape of the analytical IMF plot. The result was taken as a useful feature to distinguish normal lung sound signal with ALS. Simulation results show that the CTM of analytic IMFs has a strong ability to distinguish between normal lung sound signals and ALS.
Laila Ouannes, , Najoua Essoukri Ben Amara
Traitement du Signal, Volume 38, pp 573-585; https://doi.org/10.18280/ts.380305

Abstract:
In the recent years, the face recognition task has attracted the attention of researchers due to its efficiency in several domains such as surveillance and access control. Unfortunately, there are multiple challenges that decrease the performance of face recognition. Partial occlusion is the most challenging one since it often causes a great lack of information. The main purpose of this paper is to prove that facial reconstruction improves the results of facial recognition compared to de-occlusion and full-face recognition in the presence of occlusion. Our objective is to achieve occluded-face recognition, de-occluded-face recognition, and reconstructed-face recognition. Regarding face reconstruction, we introduce two different methods based on Laplacian pyramid blending and CycleGANs. In order to validate our work, we perform two different feature extraction techniques: hand-crafted features and learned features exploiting the final layers of a pre-trained deep architecture model. The experimental results on the EURECOM Kinect Face Dataset (EKFD) and the IST-EURECOM Light Field Face Database (IST-EURECOM LFFD) show that the proposed face reconstruction approach, compared with the face de-occlusion and occluded-face recognition ones, clearly improves the face recognition task. Our method boosts the classification performance in comparison with the state-of-the-art methods, achieving 94.66% on EKFD and 72.35% on IST-EURECOM LFFD.
Yukun Jia, Rongtao Ding, Wei Ren, Jianfeng Shu,
Traitement du Signal, Volume 38, pp 565-572; https://doi.org/10.18280/ts.380304

Abstract:
During rehabilitation, many postoperative patients need to perform autonomous massage on time and on demand. Thus, this paper develops an individualized, intelligent, and independent rehabilitation training system for based on image feature deep learning model acupoint massage that excludes human factors. The system, which innovatively integrates massage gesture recognition with human pose recognition. It relies on the binocular depth camera Kinect DK and Google MediaPipe Holistic pipeline to collect the real-time image feature data on joints and gestures of the patient in autonomous massage. Then the system calculates the coordinates of each finger joint, and computes the human poses with VGG-16, a convolutional neural network (CNN); the calculated results are translated, and presented in a virtual reality (VR) model based on Unity 3D, aiming to guide the patient actions in autonomous massage. This is because the image feature of the gesture recognition and pose recognition is hindered, when the hand or the human is occluded by the body or other things, owing to the limited recognition range of the hardware. The experimental results show that, the proposed system could correctly recognize up to 84% of non-occluded gestures, and up to 93% of non-occluded poses; the system also exhibited a good real-time performance, a high operability, and a low cost. Facing the lack of medical staff, our system can effectively improve the life quality of patients.
Loay Hassan, Adel Saleh, Mohamed Abdel-Nasser, Osama A. Omer, Domenec Puig
Traitement du Signal, Volume 38, pp 653-661; https://doi.org/10.18280/ts.380312

Abstract:
Automated cell nuclei delineation in whole-slide imaging (WSI) is a fundamental step for many tasks like cancer cell recognition, cancer grading, and cancer subtype classification. Although numerous computational methods have been proposed for segmenting nuclei in WSI images based on image processing and deep learning, existing approaches face major challenges such as color variation due to the use of different stains, the various structures of cell nuclei, and the overlapping and clumped cell nuclei. To circumvent these challenges in this article, we propose an efficient and accurate cell nuclei segmentation method based on deep learning, in which a set of accurate individual cell nuclei segmentation models are developed to predict rough segmentation masks, and then a learnable aggregation network (LANet) is used to predict the final nuclei masks. Besides, we develop cell nuclei segmentation software (with a graphical user interface—GUI) that includes the proposed method and other deep-learning-based cell nuclei segmentation methods. A challenging WSI dataset collected from different centers and organs is used to demonstrate the efficiency of our method. The experimental results reveal that our method obtains a competitive performance compared to the existing approaches in terms of the aggregated Jaccard index (AJI=89.25%) and F1-score (F1=73.02%). The developed nuclei segmentation software can be downloaded from https://github.com/loaysh2010/Cell-Nuclei-Segmentation-GUI-Application.
Mohan Goud Kathi, Jakeer Hussain Shaik
Traitement du Signal, Volume 38, pp 681-688; https://doi.org/10.18280/ts.380315

Abstract:
One of the key challenges that the computer vision is facing is the age prediction. A well efficient CNN is selected for age prediction by performing various CNN operations by taking the categories as age 40 and above age 40. The selected CNN method obtained a training accuracy of 100% at more than 100 epochs. Hence, 100 epochs is considered for training. At this, the validation accuracy achieved is 84.9%. Three kinds of age phases with an age gap of 20,10 and 5 are used to predict the age. The normal method results in very less accuracy. Hence a hierarchical method is formulated. Under the hierarchical method, CNN is trained to estimate the age gaps in decreasing order. Hence not a single classifier, a group of classifiers are used for testing the image. From traditional method to hierarchical method, the 20 age gap accuracy increased from 27% to above 60%, ten age gap increased from 12% to above 35%, and five age gap increased from 5.5% to above 21%. To improve further, the features of the face parts are derived and combined which improves the efficiency compared to normal method, but not good accuracy as Hierarchical method. The combination of hierarchical method along with the face feature extraction method results in a considerable improvement in accuracy.
, Muhammed Fatih Talu
Traitement du Signal, Volume 38, pp 619-627; https://doi.org/10.18280/ts.380309

Abstract:
Pneumonia, featured by inflammation of the air sacs in one or both lungs, is usually detected by examining chest X-ray images. This paper probes into the classification models that can distinguish between normal and pneumonia images. As is known, trained networks like AlexNet and GoogleNet are deep network architectures, which are widely adopted to solve many classification problems. They have been adapted to the target datasets, and employed to classify new data generated through transfer learning. However, the classical architectures are not accurate enough for the diagnosis of pneumonia. Therefore, this paper designs a capsule network with high discrimination capability, and trains the network on Kaggle’s online pneumonia dataset, which contains chest X-ray images of many adults and children. The original dataset consists of 1,583 normal images, and 4,273 pneumonia images. Then, two data augmentation approaches were applied to the dataset, and their effects on classification accuracy were compared in details. The model parameters were optimized through five different experiments. The results show that the highest classification accuracy (93.91% even on small images) was achieved by the capsule network, coupled with data augmentation by generative adversarial network (GAN), using optimized parameters. This network outperformed the classical strategies.
Ahmed Abdulmaged Ismael, Muhammet Baykara
Traitement du Signal, Volume 38, pp 639-651; https://doi.org/10.18280/ts.380311

Abstract:
Recently, with the explosion in the number of digital images captured every day in all life aspects, there is a growing demand for more detailed and visually attractive images. However, the images taken by current sensors are inevitably degraded by noise in various fields, such as medical, astrophysics, weather forecasting, etc., which contributes to impaired visual image quality. Therefore, work is needed to reduce noise by preserving the textural, information, and structural features of the image. So far, there are different techniques for reducing noise that various researchers have done. Each technique has its advantages and disadvantages. In this paper, a review of some significant work in the field of image denoising based on that the denoising methods is presented. These methods can be classified as wavelet domain, spatial domain, or both methods can combine to obtain the advantage them. After a brief discussion, the classification of image denoising techniques is explained. A comparative analysis of various image denoising methods is also performed to help researchers in the image denoising area. Besides, standard measurement parameters have been used to compute the results and to evaluate the performance of the used denoising techniques. This review paper aims to provide functional knowledge of image denoising methods in a nutshell for applications using images to provide ease for selecting the ideal strategy according to the necessity.
, Ramazan Tekin, Yılmaz Kaya
Traitement du Signal, Volume 38, pp 587-597; https://doi.org/10.18280/ts.380306

Abstract:
Parkinson's disease (PD) is a neurological disease that progresses further over time. Individuals suffering from this condition have a deficiency of dopamine, a neurotransmitter found in the brain's nerve cells that is critical for coordinating body movement. In this study, a new approach is proposed for the diagnosis of PD. Common Average Reference (CAR), Median Common Average Reference (MCAR), and Weighted Common Average Reference (WCAR) methods were primarily utilized to eliminate noise from the multichannel recorded walking signals in the resulting PhysioNet dataset. Statistical features were obtained from the clean walking signals following the Local Binary Pattern (LBP) transformation application. Logistic Regression (LR), Random Forest (RF), and K-nearest neighbor (Knn) methods were utilized in the classification stage. A high success rate with a value of 92.96% was observed with Knn. It was also determined that signals on which foot and the signals obtained from which point of the sole of the foot were effective in PD diagnosis in the study. In light of the findings, it was observed that noise reduction methods increased the success rate of PD diagnosis.
Back to Top Top