• Current Optics and Photonics
• /
• 제8권4호
• /
• pp.391-398
• /
• 2024

An image reconstruction algorithm is vital for the image quality of a photonic integrated interferometric imaging (PIII) system. However, image reconstruction algorithms have limitations that always lead to degraded image reconstruction. In this paper, a novel image reconstruction algorithm based on deep learning is proposed. Firstly, the principle of optical signal transmission through the PIII system is investigated. A dataset suitable for image reconstruction of the PIII system is constructed. Key aspects such as model and loss functions are compared and constructed to solve the problem of image blurring and noise influence. By comparing it with other algorithms, the proposed algorithm is verified to have good reconstruction results not only qualitatively but also quantitatively.

• 접착 및 계면
• /
• 제6권1호
• /
• pp.1-6
• /
• 2005

The high accurate simulation of very thin glue layers based on the finite element method is still connected to many problems which result from the necessity to construct a complicated mesh of essentially different sizes of elements. This can lead to a loss of accuracy, unstable calculations and even loss of convergence. However, the implementation of special transmission elements along the glue ling and special edge-elements in the near-edge region would lead to a dramatic decrease of number of finite elements in the mesh and thus, prevent unsatisfactory phenomena in numerical analysis and extensive computation time. The theoretical basis for such special elements is the knowledge about appropriate transmission conditions and the edge effects near the free boundary of the adhesive layer. Therefore, recently proposed so-called non-classical transmission conditions and the behavior near the free edge are investigated in the context of the single-lap tensile-shear test of adhesive technology.

• 한국자동차공학회논문집
• /
• 제19권1호
• /
• pp.66-72
• /
• 2011

Injection pressure is an important factor in filling procedure for injection molded parts. In addition, weldlines should be avoided to successfully produce injection molded parts. In this study, we optimally obtained injection molding process parameters that minimize injection pressure. Then, we determined the thickness of the part to avoid weldlines. To solve the optimization problem proposed, we employed MAPS-3D (Mold Analysis and Plastics Solution-3 Dimension), a commercial CAE tool for injection molding analysis, and PIAnO (Process Integration, Automation, and Optimization) as a commercial PIDO (Process Integration and Design Optimization) tool. We integrated MAPS-3D into PIAnO, automated the analysis and design procedure, and performed optimization by employing PQRSM (Progressive Quadratic Response Surface Method) equipped in PIAnO. We successfully obtained optimization results, which demonstrates the effectiveness of our design method.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.591-594
• /
• 2002

This paper presents a systematic method of probing channel characteristics and communication reliabilities of home power line communication network applied to the Internet accessed control of home appliances. The effects of the three performance deteriorating factors, i.e., additive noise, channel attenuation, and intersymbol interference, can be systematically measured by applying the channel probing waveform in the frequency range from 100㎑ to 450㎑. Probability of bit error is derived with the probed channel parameters of the signal attenuation, noise and signal-to-interference ratio read in the frequency domain. The agreement between the derived probability of bit ewer and the measured probability of bit error support the validity of the proposed approach of probing home power line channel characteristics. The experimental results performed with the constructed test-bed applying the Proposed channel probing method and the operation reliability measurement of the overall networked system also support the feasibility of commercially deploying the PLC modem installed home appliances and their services for the Internet accessed home automation in densely populated residential apartment complexes.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권2호
• /
• pp.775-792
• /
• 2015

Person re-identification is an important and challenging task in computer vision with numerous real world applications. Despite significant progress has been made in the past few years, person re-identification remains an unsolved problem. This paper presents a novel appearance-based approach to person re-identification. The approach exploits region covariance matrix and color histograms to capture the statistical properties and chromatic information of each object. Robustness against low resolution, viewpoint changes and pose variations is achieved by a novel signature, that is, the combination of Log Covariance Matrix feature and HSV histogram (LCMH). In order to further improve re-identification performance, third-party image sets are utilized as a common reference to sufficiently represent any image set with the same type. Distinctive and reliable features for a given image set are extracted through decision boundary between the specific set and a third-party image set supervised by max-margin criteria. This method enables the usage of an existing dataset to represent new image data without time-consuming data collection and annotation. Comparisons with state-of-the-art methods carried out on benchmark datasets demonstrate promising performance of our method.

• 한국멀티미디어학회논문지
• /
• 제21권2호
• /
• pp.98-107
• /
• 2018

In this paper, we propose a method to classify the skeletal maturity with a small amount of hand wrist X-ray image using deep learning-based meta-learning. General deep-learning techniques require large amounts of data, but in many cases, these data sets are not available for practical application. Lack of learning data is usually solved through transfer learning using pre-trained models with large data sets. However, transfer learning performance may be degraded due to over fitting for unknown new task with small data, which results in poor generalization capability. In addition, medical images require high cost resources such as a professional manpower and mcuh time to obtain labeled data. Therefore, in this paper, we use meta-learning that can classify using only a small amount of new data by pre-trained models trained with various learning tasks. First, we train the meta-model by using a separate data set composed of various learning tasks. The network learns to classify the bone maturity using the bone maturity data composed of the radiographs of the wrist. Then, we compare the results of the classification using the conventional learning algorithm with the results of the meta learning by the same number of learning data sets.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권9호
• /
• pp.4307-4325
• /
• 2016

With the aim of solving the problems of long processing times, high memory consumption and low event throughput in the current processing approaches in out-of-order RFID event streams, an efficient complex event processing method based on INFA-HTS (Improved Nondeterministic Finite Automaton-Hash Table Structure) is presented in this paper. The contribution of this paper lies in the fact that we use INFA and HTS to successfully realize the detection of complex events for out-of-order RFID event streams. Specifically, in our scheme, to detect the disorder of out-of-order event streams, we expand the traditional NFA model into a new INFA model to capture the related RFID primitive events from the out-of-order event stream. To high-efficiently manage the large intermediate capturing results, we use the HTS to store and process them. As a result, these problems in the existing methods can be effectively solved by our scheme. The simulation results of our experiments show that our proposed method in this paper outperforms some of the current general processing approaches used to process out-of-order RFID event streams.

• 한국정밀공학회지
• /
• 제29권1호
• /
• pp.114-120
• /
• 2012

We present a simple and low-cost method to fabricate poly(methyl-methacrylate) (PMMA) nanochannels with various shapes by combining the standard optical lithography with a PMMA layer transfer and collapse technique. We utilized PMMA membrane reflowing/collapsing phenomena into microchannels to fabricate nanochannels at both corners of arbitrarily-shaped microchannels. This allows nanochannels with various shapes such as curved nanochannels as well as straight nanochannels to be easily fabricated since the shape of the microchannel determines the shape of the nanochannels. This nanochannel fabrication method is simple, flexible, and low-cost since the standard optical lithography with low-resolution optical masks can be used to fabricate nanoscale channels as small as 100 nm wide with various shapes. Also, the sealing of nanochannels can be naturally achieved while the nanochannels are formed through the polymer layer transfer and collapse.

• 한국정밀공학회지
• /
• 제19권6호
• /
• pp.128-135
• /
• 2002

Burr is a projected part of finished workpiece. It is unavoidable and undesirable by-product of most metal cutting or shearing process. Also, it must be removed to improve the fit of machined parts, safety of workers, and the effectiveness of finishing operation. But deburring process Is one of manufacturing processes that have not been successfully automated, so deburring automation is strongly needed. This paper focused on developing a basic algorithm to find edge of workpiece and match two different image data for deburring automation which includes automatic recognition of parts, generation of deburring tool paths and edge/comer finding ability by analyzing the DXF drawing file which contains information of part geometry. As an algorithm fur corner finding, SUSAN method was chosen. It makes good performance in finding edge and corner in suitable time. And this paper suggested a simple algorithm to find matching point between CCD image and drawing file.

• 한국정밀공학회지
• /
• 제19권5호
• /
• pp.139-148
• /
• 2002

Motion errors of linear motion guideway are analyzed theoretically in this paper. For the analysis, an new algorithm predicting motion errors of bearing and guideway is proposed using the Hertz's elastic deformation theory. Accuracy averaging effect can be calculated quantitatively by analyzing relationship between motion errors of guideway and spatial frequency of rail form error. Influences of design parameters on the motion errors including the number of balls, preload, ball diameter, bearing length and the number of bearings are analyzed. As it is difficult to measure the rail form error, experimental results are compared with results analyzed by the equivalent analysis method which evaluate the motion errors of guideway using the measured errors of bearing. From the experimental results, it is confirmed that the proposed analysis method it effective lo analyze the motion errors of linear motion bearing and guideway.