Search Results (4)

Terahertz (THz) nondestructive testing (NDT) technology has been increasingly applied to the internal defect detection of composite materials. However, the THz image is affected by background noise and power limitation, leading to poor THz image quality. The recognition rate based on traditional machine vision algorithms is not high. The above methods are usually unable to determine surface defects in a timely and accurate manner. In this paper, we propose a method to detect the internal defects of composite materials by using terahertz images based on a faster region-convolutional neural networks (faster R-CNNs) algorithm. Terahertz images showing internal defects in composite materials are first acquired by a terahertz time-domain spectroscopy system. Then the terahertz images are filtered, the blurred images are removed, and the remaining images are enhanced with data and annotated with image defects to create a dataset consistent with the internal defects of the material. On the basis of the above work, an improved faster R-CNN algorithm is proposed in this paper. The network can detect various defects in THz images by changing the backbone network, optimising the training parameters, and improving the prior box algorithm to improve the detection accuracy and efficiency of the network. By taking the commonly used composite sandwich structure as a representative, a sample with typical defects is designed, and the image data are obtained through the test. Comparing the proposed method with other existing network methods, the former proves to have the advantages of a short training time and high detection accuracy. The results show that the mean average precision (mAP) without data enhancement reached 95.50%, and the mAP with data enhancement reached 98.35% and exceeded the error rate of human eye detection (5%). Compared with the original faster R-CNN algorithm of 84.39% and 85.12%, the improvement is 11.11% and 10.23%, respectively, which demonstrates superb feature extraction capability and reduces the occurrence of network errors and omissions. Full article

Terahertz (THz) spectroscopy is the de facto method to study the vibration modes and rotational energy levels of molecules and is a widely used molecular sensor for non-destructive inspection. Here, based on the THz spectra of 20 amino acids, a method that extracts high-dimensional features from a hybrid spectrum combined with absorption rate and refractive index is proposed. A convolutional neural network (CNN) calibrated by efficient channel attention (ECA) is designed to learn from the high-dimensional features and make classifications. The proposed method achieves an accuracy of $99.9\%$ and $99.2\%$ on two testing datasets, which are $12.5\%$ and $23\%$ higher than the method solely classifying the absorption spectrum. The proposed method also realizes a processing speed of 3782.46 frames per second (fps), which is the highest among all the methods in comparison. Due to the compact size, high accuracy, and high speed, the proposed method is viable for future applications in THz chemical sensors. Full article

Microfluidic devices have been extensively investigated in recent years in fields including ligand-binding analysis, chromatographic separation, molecular dynamics, and DNA sequencing. To prolong the observation of a single molecule in aqueous buffer, the solution in a sub-micron scale channel is driven by a electric field and reversed after a fixed delay following each passage, so that the molecule passes back and forth through the laser focus and the time before irreversible photobleaching is extended. However, this practice requires complex chemical treatment to the inner surface of the channel to prevent unexpected sticking to the surface and the confined space renders features, such as a higher viscosity and lower dielectric constant, which slow the Brownian motion of the molecule compared to the bulk solution. Additionally, electron beam lithography used for the fabrication of the nanochannel substantially increases the cost, and the sub-micron dimensions make the molecule difficult to locate. In this paper, we propose a method of single-molecule recycling in a capillary microchannel. A commercial fused-silica capillary with an inner diameter of 2 microns is chopped into a 1-inch piece and is fixed onto a cover slip. Two o-rings on the sides used as reservoirs and an o-ring in the middle used as observation window are glued over the capillary. The inner surface of the capillary is chemically processed to reduce the non-specific sticking and to improve capillary effect. The device does not require high-precision fabrication and thus is less costly and easier to prepare than the nanochannel. 40 nm Fluospheres^® in 50% methanol are used as working solution. The capillary is translated by a piezo stage to recycle the molecule, which diffuses freely through the capillary, and a confocal microscope is used for fluorescence collection. The passing times of the molecule through the laser focus are calculated by a real-time control system based on an FPGA, and the commands of translation are given to the piezo stage through a feedback algorithm. The larger dimensions of the capillary overcomes the strong sticking, the reduced diffusivity, and the difficulty of localizing the molecule. We have achieved a maximum number of recycles of more than 200 and developed a maximum-likelihood estimation of the diffusivity of the molecule, which attains results of the same magnitude as the previous report. This technique simplifies the overall procedure of the single-molecule recycling and could be useful for the ligand-binding studies in high-throughput screening. Full article

The taxonomy of Australian Isoodon bandicoots has changed continuously over the last 20 years, with recent genetic studies indicating discordance of phylogeographic units with current taxonomic boundaries. Uncertainty over species relationships within southern and western Isoodon, encompassing I. obesulus, I. auratus, and I. fusciventer, has been ongoing and hampered by limited sampling in studies to date. Identification of taxonomic units remains a high priority, as all are threatened to varying extents by ongoing habitat loss and feral predation. To aid diagnosis of conservation units, we increased representative sampling of I. auratus and I. fusciventer from Western Australia (WA) and investigated genetic relationships of these with I. obesulus from South Australia (SA) and Victoria (Vic) using microsatellite markers and mitochondrial DNA. mtDNA analysis identified three major clades concordant with I. obesulus (Vic), I. auratus, and I. fusciventer; however, I. obesulus from SA was polyphyletic to WA taxa, complicating taxonomic inference. Microsatellite data aided identification of evolutionarily significant units consistent with existing taxonomy, with the exception of SA I. obesulus. Further, analyses indicated SA and Vic I. obesulus have low diversity, and these populations may require more conservation efforts than others to reduce further loss of genetic diversity. Full article