Deep learning-based obstacle-avoiding autonomous UAVs with fiducial marker-based localization for structural health monitoring.

This paper proposes a framework for obstacle-avoiding autonomous unmanned aerial vehicle (UAV) systems with a new obstacle avoidance method (OAM) and localization method for autonomous UAVs for structural health monitoring (SHM) in GPS-denied areas. There are high possibilities of obstacles in the planned trajectory of autonomous UAVs used for monitoring purposes. A traditional UAV localization method with an ultrasonic beacon is limited to the scope of the monitoring and vulnerable to both depleted battery and environmental electromagnetic fields. To overcome these critical problems, a deep learning-based OAM with the integration of You Only Look Once version 3 (YOLOv3) and a fiducial marker-based UAV localization method are proposed. These new obstacle avoidance and localization methods are integrated with a real-time damage segmentation method as an autonomous UAV system for SHM. In indoor testing and outdoor tests in a large parking structure, the proposed methods showed superior performances in obstacle avoidance and UAV localization compared to traditional approaches.

Dual encoder-decoder-based deep polyp segmentation network for colonoscopy images.

Detection of colorectal polyps through colonoscopy is an essential practice in prevention of colorectal cancers. However, the method itself is labor intensive and is subject to human error. With the advent of deep learning-based methodologies, and specifically convolutional neural networks, an opportunity to improve upon the prognosis of potential patients suffering with colorectal cancer has appeared with automated detection and segmentation of polyps. Polyp segmentation is subject to a number of problems such as model overfitting and generalization, poor definition of boundary pixels, as well as the model's ability to capture the practical range in textures, sizes, and colors. In an effort to address these challenges, we propose a dual encoder-decoder solution named Polyp Segmentation Network (PSNet). Both the dual encoder and decoder were developed by the comprehensive combination of a variety of deep learning modules, including the PS encoder, transformer encoder, PS decoder, enhanced dilated transformer decoder, partial decoder, and merge module. PSNet outperforms state-of-the-art results through an extensive comparative study against 5 existing polyp datasets with respect to both mDice and mIoU at 0.863 and 0.797, respectively. With our new modified polyp dataset we obtain an mDice and mIoU of 0.941 and 0.897 respectively.

Efficient attention-based deep encoder and decoder for automatic crack segmentation.

Recently, crack segmentation studies have been investigated using deep convolutional neural networks. However, significant deficiencies remain in the preparation of ground truth data, consideration of complex scenes, development of an object-specific network for crack segmentation, and use of an evaluation method, among other issues. In this paper, a novel semantic transformer representation network (STRNet) is developed for crack segmentation at the pixel level in complex scenes in a real-time manner. STRNet is composed of a squeeze and excitation attention-based encoder, a multi head attention-based decoder, coarse upsampling, a focal-Tversky loss function, and a learnable swish activation function to design the network concisely by keeping its fast-processing speed. A method for evaluating the level of complexity of image scenes was also proposed. The proposed network is trained with 1203 images with further extensive synthesis-based augmentation, and it is investigated with 545 testing images (1280 × 720, 1024 × 512); it achieves 91.7%, 92.7%, 92.2%, and 92.6% in terms of precision, recall, F1 score, and mIoU (mean intersection over union), respectively. Its performance is compared with those of recently developed advanced networks (Attention U-net, CrackSegNet, Deeplab V3+, FPHBN, and Unet++), with STRNet showing the best performance in the evaluation metrics-it achieves the fastest processing at 49.2 frames per second.

Semi-Automated Air-Coupled Impact-Echo Method for Large-Scale Parkade Structure.

Structural Health Monitoring (SHM) has moved to data-dense systems, utilizing numerous sensor types to monitor infrastructure, such as bridges and dams, more regularly. One of the issues faced in this endeavour is the scale of the inspected structures and the time it takes to carry out testing. Installing automated systems that can provide measurements in a timely manner is one way of overcoming these obstacles. This study proposes an Artificial Neural Network (ANN) application that determines intact and damaged locations from a small training sample of impact-echo data, using air-coupled microphones from a reinforced concrete beam in lab conditions and data collected from a field experiment in a parking garage. The impact-echo testing in the field is carried out in a semi-autonomous manner to expedite the front end of the in situ damage detection testing. The use of an ANN removes the need for a user-defined cutoff value for the classification of intact and damaged locations when a least-square distance approach is used. It is postulated that this may contribute significantly to testing time reduction when monitoring large-scale civil Reinforced Concrete (RC) structures.

Field Measurement-Based System Identification and Dynamic Response Prediction of a Unique MIT Building.

Tall buildings are ubiquitous in major cities and house the homes and workplaces of many individuals. However, relatively few studies have been carried out to study the dynamic characteristics of tall buildings based on field measurements. In this paper, the dynamic behavior of the Green Building, a unique 21-story tall structure located on the campus of the Massachusetts Institute of Technology (MIT, Cambridge, MA, USA), was characterized and modeled as a simplified lumped-mass beam model (SLMM), using data from a network of accelerometers. The accelerometer network was used to record structural responses due to ambient vibrations, blast loading, and the October 16th 2012 earthquake near Hollis Center (ME, USA). Spectral and signal coherence analysis of the collected data was used to identify natural frequencies, modes, foundation rocking behavior, and structural asymmetries. A relation between foundation rocking and structural natural frequencies was also found. Natural frequencies and structural acceleration from the field measurements were compared with those predicted by the SLMM which was updated by inverse solving based on advanced multiobjective optimization methods using the measured structural responses and found to have good agreement.