Search

Abstract Wild bird repulsion is critical in agriculture because it helps avoid agricultural food losses and mitigates the risk of avian influenza. Wild birds transmit avian influenza in poultry farms and thus cause large economic losses. In this study, we developed an automatic wild bird repellent system that is based on deep-learning-based wild bird detection and integrated with a laser rotation mechanism. When a wild bird appears at a farm, the proposed system detects the bird’s position in an image captured by its detection unit and then uses a laser beam to repel the bird. The wild bird detection model of the proposed system was optimized for detecting small pixel targets, and trained through a deep learning method by using wild bird images captured at different farms. Various wild bird repulsion experiments were conducted using the proposed system at an outdoor duck farm in Yunlin, Taiwan. The statistical test results of our experimental data indicated that the proposed automatic wild bird repellent system effectively reduced the number of wild birds in the farm. The experimental results indicated that the developed system effectively repelled wild birds, with a high repulsion rate of 40.3% each day.

Creating a personalized implant for cranioplasty can be costly and aesthetically challenging, particularly for comminuted fractures that affect a wide area. Despite significant advances in deep learning techniques for 2D image completion, generating a 3D shape inpainting remains challenging due to the higher dimensionality and computational demands for 3D skull models. Here, we present a practical deep-learning approach to generate implant geometry from defective 3D skull models created from CT scans. Our proposed 3D reconstruction system comprises two neural networks that produce high-quality implant models suitable for clinical use while reducing training time. The first network repairs low-resolution defective models, while the second network enhances the volumetric resolution of the repaired model. We have tested our method in simulations and real-life surgical practices, producing implants that fit naturally and precisely match defect boundaries, particularly for skull defects above the Frankfort horizontal plane.

This article has been retracted. Please see the Retraction Notice for more detail: https://doi.org/10.1186/s12950-015-0074-3.

Abstract We present a 3D deep learning framework that can generate a complete cranial model using a defective one. The Boolean subtraction between these two models generates the geometry of the implant required for surgical reconstruction. There is little or no need for post-processing to eliminate noise in the implant model generated by the proposed approach. The framework can be used to meet the repair needs of cranial imperfections caused by trauma, congenital defects, plastic surgery, or tumor resection. Traditional implant design methods for skull reconstruction rely on the mirror operation. However, these approaches have great limitations when the defect crosses the plane of symmetry or the patient's skull is asymmetrical. The proposed deep learning framework is based on an enhanced three-dimensional autoencoder. Each training sample for the framework is a pair consisting of a cranial model converted from CT images and a corresponding model with simulated defects on it. Our approach can learn the spatial distribution of the upper part of normal cranial bones and use flawed cranial data to predict its complete geometry. Empirical research on simulated defects and actual clinical applications shows that our framework can meet most of the requirements of cranioplasty.

Aggressive tracing of contacts of confirmed cases is crucial to Taiwan's successful control of the early spread of COVID-19. As the pandemic lingers, an epidemiological investigation that can be conducted efficiently in a timely manner can help decrease the burden on the health personnel and increase the usefulness of such information in decision making. To develop a new tool that can improve the current practice of epidemiological investigation by incorporating new technologies in digital platform and knowledge graphs. To meet the various needs of the epidemiological investigation, we decided to develop an e-Outbreak Platform that provides a semi-structured, multifaceted, computer-aided questionnaire for outbreak investigation. There are three major parts of the platform: (1) a graphic portal that allows users to have an at-glance grasp of the functions provided by the platform and then choose the one they need; (2) disease-specific questionnaires that can accommodate different formats of the information, including text typing, button selection, and pull-down menu; and (3) functions to utilize the stored information, including report generation, statistical analyses, and knowledge graphs displaying contact tracing. When the number of outbreak investigation increases, the knowledge graphs can be extended to encompass other persons appearing in the same location at the same time, i.e., constituting a potential contact cluster. The information extracted can also be used to display the tracing on a map in animation. Overall, this system can provide a basis for further refinement that can be generalized to a variety of outbreak investigations.

Applying edge preservation filters for cost aggregation has been a leading technique in generating dense disparity maps. However, traditional approaches usually require intensive calculations, and their design parameters must be tuned for different scenarios to obtain the best performance. This paper shows that a simple texture-independent aggregation approach can achieve similar high performance. The proposed algorithm is equivalent to a sequence of matrix multiplications involving two weighting matrices and a primary matching cost. Notably, the weighting matrices are constant for image pairs with the same resolution. For higher matching accuracy, we integrate the algorithm with a multi-scale scheme to fully exploit the spatial distribution of textures in the image pairs. The resultant hybrid approach is efficient and accurate enough to surpass most existing approaches in stereo matching. The performance of the proposed approach is verified by extensive simulation results using the Middlebury (3rd Edition) benchmark stereo database.

Southwest (SW) Taiwan lies on the deformation front of a plate collision zone. Most earthquakes in this region occur due to deformation of the upper crust, therefore, potential seismic hazards in this area must be considered. On 6 February 2016, a moderate, but disastrous earthquake (ML = 6.6; depth = 14.6 km) occurred in Meinong, a district in the Kaohsiung area of Taiwan. This earthquake produced three cluster-like aftershocks. Among them, two of the aftershocks clusters were not spatially correlated with the main earthquake. Additionally, the trends of these two seismic clusters were not associated with previously known faults or geological structures. Therefore, our major intention is not only to investigate the rupture process of the 2016 Meinong earthquake, but more importantly, to look into the corresponding seismogenic process. In this research, high quality strong-motion data was used to invert the slip distribution on the fault plane using the isochron method. In addition, we relocated aftershocks to further obtain focal mechanisms using a 3D velocity model. Coulomb failure stress maps are calculated for different depth ranges based on the obtained source slip distribution to verify whether the main-shock triggers nearby unknown faults or not. In conclusion, we suggest that there might exist a NW-SE trending, north-dipping fault, which is located north of the source area and may have been triggered by the initial shock. We also conclude that the 2016 Meinong earthquake did trigger the pre-existing normal faults beneath Tainan City.

The vigorous collision between the Eurasian plate and Philippine Sea plate in Taiwan causes a series of imbricate fold and thrust belts to develop at the deformation front. The Chukou Fault (CKF), characterized by a thrust type fault, located in Chiayi County, southwest (SW) Taiwan, is a prominent boundary between the fold-thrust belts and the Western Coastal Plain. Most of the seismicity in SW Taiwan is associated with this fault and its neighboring fault systems. The seismotectonic structures in the CKF zone, especially in the east, are complex due to the interactions among fault systems with distinct slip motions. To gain better insights into the seismogenic characteristics in the CKF zone, we used 1661 microearthquakes recorded by a temporary dense broadband seismic network and the Central Weather Bureau Seismic Network (CWBSN) between 2003 and 2004 to investigate the physical properties of the crust in the CKF zone. A waveform cross-correlation technique was applied to 143086 pairs of waveform data to determine the relative differential travel time between the P- and S-waves. By combining both the absolute and relative differential travel time data, we were able to obtain a new 3-D crustal P-wave velocity structure and Vp/Vs ratios. This study suggests that by using both absolute and relative differential travel time data in tomographic inversion can obtain precise 3-D velocity images and also gain better correlation between seismic events and fault structures, which is crucial for understanding the seismogenic process in our study area.

Movable suspended microstructures are the common feature of sensors or devices in the fields of Complementary-Metal-Oxide-Semiconductors and Micro-Electro-Mechanical Systems which are usually abbreviated as CMOS-MEMS. To suspend the microstructures, it is commonly to etch the sacrificial layer under the microstructure layer. For large-area microstructures, it is necessary to design a large number of etching holes on the microstructure to enhance the etchant uniformly and rapidly permeate into the sacrificial layer. This paper aims at evaluating the fringe capacitance caused by etching holes on microstructures and developing empirical formulas. The formula of capacitance compensation term is derived by curve-fitting on the simulation results by the commercial software ANSYS. Compared with the ANSYS simulation, the deviation of the present formula is within ±5%. The application to determine the capacitance of an electrostatic micro-beam with etching holes is demonstrated in a microstructure experiment, which agrees very well with the experimental data, and the maximum deviation is within ±8%. The present formula is with simple form, wide application range, high accuracy, and easy to use. It is expected to provide the micro-device designers to estimate the capacitance of microstructures with etching holes and predominate in the device characteristics.

Abstract Background High-mobility group box 1 (HMGB1) was observed to be an important extracellular mediator involved in vascular inflammation associated with subarachnoid hemorrhage (SAH). This study is of interest to examine the efficacy of 4′-O-β-d-glucosyl-5-O-methylvisamminol (4OGOMV), C22H28O10, on the alternation of cytokines and HMGB1 in an animal model. Methods A rodent double hemorrhage SAH model was employed. Administration with 4OGOMV was initiated 1 h after animals were subjected to SAH. Basilar arteries (BAs) were harvested and cortexes examined for HMGB1 mRNA, protein expression (Western blot) and monocyte chemoattractant protein-1 (MCP-1) immunostaining. Cerebrospinal fluid samples were collected to examine IL-1β, IL-6, IL-8 and MCP-1 (rt-PCR). Results Morphological findings revealed endothelial cell deformity, intravascular elastic lamina torture, and smooth muscle necrosis in the vessels of SAH groups. Correspondently, IL-1β, IL-6 and MCP-1 in the SAH-only and SAH-plus vehicle groups was also elevated. 4OGOMV dose-dependently reduced HMGB1 protein expression when compared with the SAH groups.(p

Abstract Background Up-regulation of regulated upon activation, normal T-cell expressed and secreted (RANTES/CCL5) and adhesion molecules is observed in the serum of animals following experimental subarachnoid hemorrhage (SAH). The present study was to examine the effect of valproic acid (VPA) on RANTES and alternation of adhesion molecules in this model. Methods A rodent SAH model was employed. Animals were randomly assigned into six groups. Basilar artery (BA) was harvested for intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule–1 (VCAM-1), and E-selectin evaluation (western blotting) and RANTES (rt-PCR). 1 ng CCL5 recombinant protein intrathecal injection was performed in the VPA + SAH groups. (N = 5). Results Convoluted internal elastic lamina, distorted endothelial wall, and smooth muscle micro-necrosis was prominently observed in the SAH groups, which is absent in the VPA treatment and the healthy controls. Treatment with VPA dose-dependently reduced the ICAM-1, E-selectin and RANTES level, compared with the SAH group (p

The high quality properties and benefits of graphene-oxide have generated an active area of research where many investigations have shown potential applications in various technological fields. This paper proposes a methodology for enhancing the pyro-electricity of PVDF by graphene-oxide doping. The PVDF film with graphene-oxide is prepared by the sol-gel method. Firstly, PVDF and graphene-oxide powders are dispersed into dimethylformamide as solvent to form a sol solution. Secondly, the sol solution is deposited on a flexible ITO/PET substrate by spin-coating. Thirdly, the particles in the sol solution are polymerized through baking off the solvent to produce a gel in a state of a continuous network of PVDF and graphene-oxide. The final annealing process pyrolyzes the gel and form a β-phase PVDF film with graphene-oxide doping. A complete study on the process of the graphene oxide doping of PVDF is accomplished. Some key points about the process are addressed based on experiments. The solutions to some key issues are found in this work, such as the porosity of film, the annealing temperature limitation by the use of flexible PET substrate, and the concentrations of PVDF and graphene-oxide.

Stereo matching is complicated by the uneven distribution of textures on the image pairs. We address this problem by applying the edge-preserving guided-Image-filtering (GIF) at different resolutions. In contrast to most multi-scale stereo matching algorithms, parameters of the proposed hierarchical GIF model are in an innovative weighted-combination scheme to generate an improved matching cost volume. Our method draws its strength from exploiting texture in various resolution levels and performing an effective mixture of the derived parameters. This novel approach advances our recently proposed algorithm, the pervasive guided-image-filtering scheme, by equipping it with hierarchical filtering modules, leading to disparity images with more details. The approach ensures as many different-scale patterns as possible to be involved in the cost aggregation and hence improves matching accuracy. The experimental results show that the proposed scheme achieves the best matching accuracy when compared with six well-recognized cutting-edge algorithms using version 3 of the Middlebury stereo evaluation data sets.

The internal temperature is an important index for the prevention and maintenance of a spindle. However, the temperature inside the spindle is undetectable directly because there is no space to embed a temperature sensor, and drilling holes will reduce its mechanical stiffness. Therefore, it is worthwhile understanding the thermal-feature of a spindle. This article presents a methodology to identify the thermal-feature model of an externally driven spindle. The methodology contains self-made hardware of the temperature sensing and wireless transmission module (TSWTM) and software for the system identification (SID); the TSWTM acquires the temperature training data, while the SID identifies the parameters of the thermal-feature model of the spindle. Then the resulting thermal-feature model is written into the firmware of the TSWTM to give it the capability of accurately calculating the internal temperature of the spindle from its surface temperature during the operation, or predicting its temperature at various speeds. The thermal-feature of the externally driven spindle is modeled by a linearly time-invariant state-space model whose parameters are identified by the SID, which integrates the command “n4sid” provided by the System ID Toolbox of MATLAB and the k-fold cross-validation that is common in machine learning. The present SID can effectively strike a balance between the bias and variance of the model, such that both under-fitting and over-fitting can be avoided. The resulting thermal-feature model can not only predict the temperature of the spindle rotating at various speeds but can also calculate the internal temperature of the spindle from its surface temperature. Its validation accuracy is higher than 98.5%. This article illustrates the feasibility of accurately calculating the internal temperature (undetectable directly) of the spindle from its surface temperature (detectable directly).