• Journal of the Semiconductor & Display Technology
• /
• v.20 no.3
• /
• pp.6-10
• /
• 2021

Recognition of characters printed or marked on the PCB components from images captured using cameras is an important task in PCB components inspection systems. Previous optical character recognition (OCR) of PCB components typically consists of two stages: character segmentation and classification of each segmented character. However, character segmentation often fails due to corrupted characters, low image contrast, etc. Thus, OCR without character segmentation is desirable and increasingly used via deep neural networks. Typical implementation based on deep neural nets without character segmentation includes convolutional neural network followed by recurrent neural network (RNN). However, one disadvantage of this approach is slow execution due to RNN layers. LPRNet is a segmentation-free character recognition network with excellent accuracy proved in license plate recognition. LPRNet uses a wide convolution instead of RNN, thus enabling fast inference. In this paper, LPRNet was adapted for recognizing characters printed on PCB components with fast execution and high accuracy. Initial training with synthetic images followed by fine-tuning on real text images yielded accurate recognition. This net can be further optimized on Intel CPU using OpenVINO tool kit. The optimized version of the network can be run in real-time faster than even GPU.

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.3
• /
• pp.113-116
• /
• 2021

As artificial intelligence technologies, including deep learning, develop, these technologies are being introduced to code similarity analysis. In the traditional analysis method of calculating the graph edit distance (GED) after converting the source code into a control flow graph (CFG), there are studies that calculate the GED through a trained graph neural network (GNN) with the converted CFG, Methods for analyzing code similarity through CNN by imaging CFG are also being studied. In this paper, to determine which approach will be effective and efficient in researching code similarity analysis methods using artificial intelligence in the future, code similarity is measured through funcGNN, which measures code similarity using GNN, and Siamese Network, which is an image similarity analysis model. The accuracy was compared and analyzed. As a result of the analysis, the error rate (0.0458) of the Siamese network was bigger than that of the funcGNN (0.0362).

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.1
• /
• pp.99-103
• /
• 2021

Warpage defect detecting of scaffold is very important in biosensor production. Because warpaged scaffold cause problem in cell culture. Currently, there is no detection equipment to warpaged scaffold. In this paper, we produced detection model for shape warpage detection using deep learning based CNN. We confirmed the shape of the scaffold that is widely used in cell culture. We produced scaffold specimens, which are widely used in biosensor fabrications. Then, the scaffold specimens were photographed to collect image data necessary for model manufacturing. We produced the detecting model of scaffold warpage defect using Densenet among CNN models. We evaluated the accuracy of the defect detection model with mAP, which evaluates the detection accuracy of deep learning. As a result of model evaluating, it was confirmed that the defect detection accuracy of the scaffold was more than 95%.

• Journal of the Semiconductor & Display Technology
• /
• v.21 no.2
• /
• pp.117-122
• /
• 2022

This paper proposes a method for evaluating the work of manufacturing workers using MediaPipe as a risk factor for musculoskeletal diseases. Recently, musculoskeletal disorders (MSDs) caused by repeated working attitudes in industrial sites have emerged as one of the biggest problems in the industrial health field while increasing public interest. The Korea Occupational Safety and Health Agency presents tools such as NIOSH Lifting Equations (NIOSH), OWAS (Ovako Working-posture Analysis System), Rapid Upper Limb Assessment (RULA), and Rapid Entertainment Assessment (REBA) as ways to quantitatively calculate the risk of musculoskeletal diseases that can occur due to workers' repeated working attitudes. To compensate for these shortcomings, the system proposed in this study obtains the position of the joint by estimating the posture of the worker using the posture estimation learning model of MediaPipe. The position of the joint is calculated using inverse kinetics to obtain an angle and substitute it into the REBA equation to calculate the load level of the working posture. The calculated result was compared to the expert's image-based REBA evaluation result, and if there was a result with a large error, feedback was conducted with the expert again.

• Journal of Sensor Science and Technology
• /
• v.32 no.6
• /
• pp.425-431
• /
• 2023

An asymmetric metal-semiconductor-metal Al_0.24Ga_0.76N ultraviolet (UV) sensor was fabricated, and the effects of local Joule heating were investigated. After dielectric breakdown, the current density under a reverse bias of 2.0 V was 1.1×10^-9 A/cm², significantly lower than 1.2×10^-8 A/cm² before dielectric breakdown; moreover, the Schottky behavior of the Ti/Al/Ni/Au electrode changed to ohmic behavior under forward bias. The UV-to-visible rejection ratio (UVRR) under a reverse bias of 7.0 V before dielectric breakdown was 87; however, this UVRR significantly increased to 578, in addition to providing highly reliable responsivity. Transmission electron microscopy revealed interdiffusion between adjacent layers, with nitrogen vacancies possibly formed owing to local Joule heating at the AlGaN/Ti/Al/Ni/Au interfaces. X-ray photoelectron microscopy results revealed decreases in the peak intensities of the O 1s binding energies associated with the Ga-O bond and OH-, which act as electron-trapping states on the AlGaN surface. The reduction in dark current owing to the proposed local heating method is expected to increase the sensing performance of UV optoelectronic integrated devices, such as active-pixel UV image sensors.

• Journal of the Semiconductor & Display Technology
• /
• v.22 no.4
• /
• pp.119-123
• /
• 2023

This study is normal orientation technology of slit coating equipment to improve the quality of curved displays. Currently, the demand for curved displays is increasing significantly due to advantages such as screen immersion or design in various industries. Accordingly, changes in the display coating process are essential. In the curved display coating process, unlike the existing flat coating process, the nozzle must be rotated along the curvature of the curved surface to spray the coating solution. The coating solution must be applied while maintaining a uniform thickness. If the thickness of the coating liquid applied to the target surface is non-uniform, the quality of the product may be degraded such as image quality deterioration and light spreading. This paper presents technology and experimental results for keeping the nozzle of slit coating equipment perpendicular to the curved surface and is expected to contribute to the quality improvement of curved displays.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.17 no.3
• /
• pp.446-457
• /
• 2017

In order to reduce the size of frame memory or bus bandwidth, frame memory compression (FMC) recompresses reconstructed or reference frames of video codecs. This paper proposes a novel FMC design based on discrete wavelet transform (DWT) - set partitioning in hierarchical trees (SPIHT), which supports fine-scalable throughput and is area-efficient. In the proposed design, multi-cores with small block sizes are used in parallel instead of a single core with a large block size. In addition, an appropriate pipelining schedule is proposed. Compared to the previous design, the proposed design achieves the processing speed which is closer to the target system speed, and therefore it is more efficient in hardware utilization. In addition, a scheme in which two passes of SPIHT are merged into one pass called merged refinement pass (MRP) is proposed. As the number of shifters decreases and the bit-width of remained shifters is reduced, the size of SPIHT hardware significantly decreases. The proposed FMC encoder and decoder designs achieve the throughputs of 4,448 and 4,000 Mpixels/s, respectively, and their gate counts are 76.5K and 107.8K. When the proposed design is applied to high efficiency video codec (HEVC), it achieves 1.96% lower average BDBR and 0.05 dB higher average BDPSNR than the previous FMC design.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.7
• /
• pp.83-89
• /
• 1989

Ferroelectric $PbTiO_3$ thin film is deposited with rf sputtering at substrate temperature of $100-150^{\circ}C$. It is found that this has pyrochlore structure of amorphous type by X-ray diffractive analysis. Thermal annealing has excellent characteristics at $550^{\circ}C$ and laser annealing has best crystalline structure in case of scanning with 50 watts. Interface states in MFST and MFOST structure with a $PbTiO_3$ ferroelectric thin film gate have been investigated from analysis of C-V data. The interface states density has been drastically reduced by inserting an oxide layer between ferroelectric and semiconductor. The observed effect increase feasibility of employing ferroelectric thin films such as nonvolatile memory field effect transistor, IR optical FET, and Image Devices with a ferroelectric layer.

• Journal of IKEEE
• /
• v.24 no.3
• /
• pp.883-889
• /
• 2020

In this paper, we propose a method to use PV(Park's Vector) pattern for inductive motor stator fault diagnosis using CNN(Convolution Neural Network). The conventional CNN based fault diagnosis method was performed by imaging three-phase currents, but this method was troublesome to perform normalization by artificially setting the starting point and phase of current. However, when using PV pattern, the problem of normalization could be solved because the 3-phase current shows a certain circular pattern. In addition, the proposed method is proved to be superior in the accuracy of CNN by 18.18[%] compared to the previous current data image due to the autonomic normalization.

• Journal of Radiation Industry
• /
• v.8 no.3
• /
• pp.141-146
• /
• 2014

Nickel (Ni) electroplating was implemented by using a metal Ni powder in order to establish a $^{63}Ni$ plating condition on the PN junction semiconductor needed for production of beta-voltaic battery. PN junction semiconductors with a Ni seed layer of 500 and $1000{\AA}$ were coated with Ni at current density from 10 to $50mA\;cm^{-2}$. The surface roughness and average grain size of Ni deposits were investigated by XRD and SEM techniques. The roughness of Ni deposit was increased as the current density was increased, and decreased as the thickness of Ni seed layer was increased. The results showed that the optimum surface shape was obtained at a current density of $10mA\;cm^{-2}$ in seed layer with thickness of $500{\AA}$, $20mA\;cm^{-2}$ of $1000{\AA}$. Also, pure Ni deposit was well coated on a PN junction semiconductor without any oxide forms. Using the line width of (111) in XRD peak, the average grain size of the Ni thick firm was measured. The results showed that the average grain size was increased as the thickness of seed layer was increased.