Autonomous Image-Based Corrosion Detection in Steel Structures Using Deep Learning.

Steel structures are susceptible to corrosion due to their exposure to the environment. Currently used non-destructive techniques require inspector involvement. Inaccessibility of the defective part may lead to unnoticed corrosion, allowing the corrosion to propagate and cause catastrophic structural failure over time. Autonomous corrosion detection is essential for mitigating these problems. This study investigated the effect of the type of encoder-decoder neural network and the training strategy that works the best to automate the segmentation of corroded pixels in visual images. Models using pre-trained DesnseNet121 and EfficientNetB7 backbones yielded 96.78% and 98.5% average pixel-level accuracy, respectively. Deeper EffiecientNetB7 performed the worst, with only 33% true-positive values, which was 58% less than ResNet34 and the original UNet. ResNet 34 successfully classified the corroded pixels, with 2.98% false positives, whereas the original UNet predicted 8.24% of the non-corroded pixels as corroded when tested on a specific set of images exclusive to the investigated training dataset. Deep networks were found to be better for transfer learning than full training, and a smaller dataset could be one of the reasons for performance degradation. Both fully trained conventional UNet and ResNet34 models were tested on some external images of different steel structures with different colors and types of corrosion, with the ResNet 34 backbone outperforming conventional UNet.

Dynamic Selection Techniques for Detecting GPS Spoofing Attacks on UAVs.

Unmanned aerial vehicles are prone to several cyber-attacks, including Global Positioning System spoofing. Several techniques have been proposed for detecting such attacks. However, the recurrence and frequent Global Positioning System spoofing incidents show a need for effective security solutions to protect unmanned aerial vehicles. In this paper, we propose two dynamic selection techniques, Metric Optimized Dynamic selector and Weighted Metric Optimized Dynamic selector, which identify the most effective classifier for the detection of such attacks. We develop a one-stage ensemble feature selection method to identify and discard the correlated and low importance features from the dataset. We implement the proposed techniques using ten machine-learning models and compare their performance in terms of four evaluation metrics: accuracy, probability of detection, probability of false alarm, probability of misdetection, and processing time. The proposed techniques dynamically choose the classifier with the best results for detecting attacks. The results indicate that the proposed dynamic techniques outperform the existing ensemble models with an accuracy of 99.6%, a probability of detection of 98.9%, a probability of false alarm of 1.56%, a probability of misdetection of 1.09%, and a processing time of 1.24 s.

A Comprehensive Survey on Spectrum Sensing in Cognitive Radio Networks: Recent Advances, New Challenges, and Future Research Directions.

Cognitive radio technology has the potential to address the shortage of available radio spectrum by enabling dynamic spectrum access. Since its introduction, researchers have been working on enabling this innovative technology in managing the radio spectrum. As a result, this research field has been progressing at a rapid pace and significant advances have been made. To help researchers stay abreast of these advances, surveys and tutorial papers are strongly needed. Therefore, in this paper, we aimed to provide an in-depth survey on the most recent advances in spectrum sensing, covering its development from its inception to its current state and beyond. In addition, we highlight the efficiency and limitations of both narrowband and wideband spectrum sensing techniques as well as the challenges involved in their implementation. TV white spaces are also discussed in this paper as the first real application of cognitive radio. Last but by no means least, we discuss future research directions. This survey paper was designed in a way to help new researchers in the field to become familiar with the concepts of spectrum sensing, compressive sensing, and machine learning, all of which are the enabling technologies of the future networks, yet to help researchers further improve the efficiently of spectrum sensing.

Wideband Spectrum Sensing: A Bayesian Compressive Sensing Approach.

Sensing the wideband spectrum is an important process for next-generation wireless communication systems. Spectrum sensing primarily aims at detecting unused spectrum holes over wide frequency bands so that secondary users can use them to meet their requirements in terms of quality-of-service. However, this sensing process requires a great deal of time, which is not acceptable for timely communications. In addition, the sensing measurements are often affected by uncertainty. In this paper, we propose an approach based on Bayesian compressive sensing to speed up the process of sensing and to handle uncertainty. This approach takes only a few measurements using a Toeplitz matrix, recovers the wideband signal from a few measurements using Bayesian compressive sensing via fast Laplace prior, and detects either the presence or absence of the primary user using an autocorrelation-based detection method. The proposed approach was implemented using GNU Radio software and Universal Software Radio Peripheral units and was tested on real-world signals. The results show that the proposed approach speeds up the sensing process by minimizing the number of samples while achieving the same performance as Nyquist-based sensing techniques regarding both the probabilities of detection and false alarm.

Automatic and robust system for correcting microarray images' rotations and isolating spots.

Microarray images contain a large volume of genetic data in the form of thousands of spots that need to be extracted and analyzed using digital image processing. Automatic extraction, gridding, is therefore necessary to save time, to remove user-dependent variations, and, hence, to obtain repeatable results. In this research paper, an algorithm that involves four steps is proposed to efficiently grid microarray images. A set of real and synthetic microarray images of different sizes and degrees of rotations is used to assess the proposed algorithm, and its efficiency is compared with the efficiencies of other methods from the literature.

An improved 1-D gel electrophoresis image analysis system.

Images obtained through the gel electrophoresis technique contain important genetic information. However, due to degradations and abnormalities from which these images suffer, extracting this information can be a tedious task and may lead to reproducibility issues. Image processing techniques that are commonly used to analyze gel electrophoresis images require three main steps: band detection, band matching, and quantification. Although several techniques were proposed to automate all steps fully, gel image analysis still requires researchers to extract information manually. This type of extraction is time consuming and subject to human errors. This paper proposes a fully automated system to analyze the gel electrophoresis images. This system involves four main steps: lane separation, lane segmentation, band detection, and data quantification.

A Deep Learning approach for Diagnosis of Mild Cognitive Impairment Based on MRI Images.

Mild cognitive impairment (MCI) is an intermediary stage condition between healthy people and Alzheimer's disease (AD) patients and other dementias. AD is a progressive and irreversible neurodegenerative disorder, which is a significant threat to people, age 65 and older. Although MCI does not always lead to AD, an early diagnosis at the stage of MCI can be very helpful in identifying people who are at risk of AD. Moreover, the early diagnosis of MCI can lead to more effective treatment, or at least, significantly delay the disease's progress, and can lead to social and financial benefits. Magnetic resonance imaging (MRI), which has become a significant tool for the diagnosis of MCI and AD, can provide neuropsychological data for analyzing the variance in brain structure and function. MCI is divided into early and late MCI (EMCI and LMCI) and sadly, there is no clear differentiation between the brain structure of healthy people and MCI patients, especially in the EMCI stage. This paper aims to use a deep learning approach, which is one of the most powerful branches of machine learning, to discriminate between healthy people and the two types of MCI groups based on MRI results. The convolutional neural network (CNN) with an efficient architecture was used to extract high-quality features from MRIs to classify people into healthy, EMCI, or LMCI groups. The MRIs of 600 individuals used in this study included 200 control normal (CN) people, 200 EMCI patients, and 200 LMCI patients. This study randomly selected 70 percent of the data to train our model and 30 percent for the test set. The results showed the best overall classification between CN and LMCI groups in the sagittal view with an accuracy of 94.54 percent. In addition, 93.96 percent and 93.00 percent accuracy were reached for the pairs of EMCI/LMCI and CN/EMCI, respectively.

Comparative analysis of various brain imaging techniques.

The bio-imaging techniques have widespread applications from diagnosing diseases to investigating the body tissues at the cells level. Traditionally, these techniques were used mainly in the orthopedic treatment. However, with the development of infrared cameras, ultrasound, and radio wave technology, they are used in different medical fields such as cardiovascular analysis, neurological treatment and infant care. This paper reviews the common bio-imaging techniques used in the brain imaging and compares them based on resolution, contrast, biological risks involved, and price.

Predicting neuropathic ulceration: analysis of static temperature distributions in thermal images.

Foot ulcers affect millions of Americans annually. Conventional methods used to assess skin integrity, including inspection and palpation, may be valuable approaches, but they usually do not detect changes in skin integrity until an ulcer has already developed. We analyze the feasibility of thermal imaging as a technique to assess the integrity of the skin and its many layers. Thermal images are analyzed using an asymmetry analysis, combined with a genetic algorithm, to examine the infrared images for early detection of foot ulcers. Preliminary results show that the proposed technique can reliably and efficiently detect inflammation and hence effectively predict potential ulceration.