Search

Low-frequency forbidden bandgap (LFB) in solid-fluid superlattice (SFSL) provides a novel degree of freedom to control the propagation of low frequency acoustic waves. However, its narrow angle range seriously restricts its practical applications. To engineer the angle range of LFB, the transmission coefficient of a cascade of multiple 1D superlattices was calculated using the transfer matrix method and also verified by finite element simulations. It was found that the LFB with ultra-wide angle range can be created using only 4 cells of a two-phased cascade structure and can be adjusted by changing the solid filling fraction of each sub-SFSL. By introducing two or more 1D superlattices into the layered cascade structure, the LFB with multiple angle ranges and/or ultra-wide angle range can be engineered for low frequency acoustic waves whose wavelengths are much larger than the lattice constant. Such acoustic low-frequency forbidden bandgap materials are useful for making perfect acoustic low-frequency filters with broad bandwidth in selective angle ranges, which are useful in subwavelength acoustic devices. [ABSTRACT FROM AUTHOR]

Detecting and locating surface defects in textured materials is a crucial but challenging problem due to factors such as texture variations and lack of adequate defective samples prior to testing. In this paper we present a novel unsupervised method for automatically detecting defects in fabrics based on a deep convolutional generative adversarial network (DCGAN). The proposed method extends the standard DCGAN, which consists of a discriminator and a generator, by introducing a new encoder component. With the assistance of this encoder, our model can reconstruct a given query image such that no defects but only normal textures will be preserved in the reconstruction. Therefore, when subtracting the reconstruction from the original image, a residual map can be created to highlight potential defective regions. Besides, our model generates a likelihood map for the image under inspection where each pixel value indicates the probability of occurrence of defects at that location. The residual map and the likelihood map are then synthesized together to form an enhanced fusion map. Typically, the fusion map exhibits uniform gray levels over defect-free regions but distinct deviations over defective areas, which can be further thresholded to produce a binarized segmentation result. Our model can be unsupervisedly trained by feeding with a set of small-sized image patches picked from a few defect-free examples. The training is divided into several successively performed stages, each under an individual training strategy. The performance of the proposed method has been extensively evaluated by a variety of real fabric samples. The experimental results in comparison with other methods demonstrate its effectiveness in fabric defect detection. [ABSTRACT FROM AUTHOR]

To speed up the object location for the IC packaging inspection, this paper develops a novel binary large object (BLOB) analysis algorithm by using run length encoding (RLE). First, the new data structures for RLE and BLOB-linked lists are designed to accelerate the access and modifications of data. Second, to avoid the traditional labeling conflicts and simplify the comparisons of connectivity branches, an efficient algorithm for BLOB analysis is presented. Furthermore, area feature of objects can be extracted when the BLOB-linked list are dynamically created or modified. Finally, to evaluate the performance of the proposed method, ICs with various types of small outline packaging packages were located by the proposed algorithm. The experimental results not only demonstrate that the IC pins could be effectively and robustly located with the proposed algorithm, but also show that the presented algorithm runs faster than other classical methods. [ABSTRACT FROM AUTHOR]