Automatic TFT-LCD Mura Inspection Based on Studentized Residuals in Regression Analysis

Yu-Chiang Chuang,Shu-Kai S. Fan 대한산업공학회 2009 Industrial Engineeering & Management Systems Vol.8 No.3

In recent days, large-sized flat-panel display (FPD) has been increasingly applied to computer monitors and TVs. Mura defects, appearing as low contrast or non-uniform brightness region, sometimes occur in manufacturing of the Thin-Film Transistor Liquid-Crystal Displays (TFT-LCD). Implementation of automatic Mura inspection methods is necessary for TFT-LCD production. Various existing Mura detection methods based on regression diagnostics, surface fitting and data transformation have been presented with good performance. This paper proposes an efficient Mura detection method that is based on a regression diagnostics using studentized residuals for automatic Mura inspection of FPD. The input image is estimated by a linear model and then the studentized residuals are calculated for filtering Mura regions. After image dilation, the proposed threshold is determined for detecting the non-uniform brightness region in TFT-LCD by means of monitoring the every pixel in the image. The experimental results obtained from several test images are used to illustrate the effectiveness and efficiency of the proposed method for Mura detection.

Automatic TFT-LCD Mura Inspection Based on Studentized Residuals in Regression Analysis

Chuang, Yu-Chiang,Fan, Shu-Kai S. Korean Institute of Industrial Engineers 2009 Industrial Engineeering & Management Systems Vol.8 No.3

In recent days, large-sized flat-panel display (FPD) has been increasingly applied to computer monitors and TVs. Mura defects, appearing as low contrast or non-uniform brightness region, sometimes occur in manufacturing of the Thin-Film Transistor Liquid-Crystal Displays (TFT-LCD). Implementation of automatic Mura inspection methods is necessary for TFT-LCD production. Various existing Mura detection methods based on regression diagnostics, surface fitting and data transformation have been presented with good performance. This paper proposes an efficient Mura detection method that is based on a regression diagnostics using studentized residuals for automatic Mura inspection of FPD. The input image is estimated by a linear model and then the studentized residuals are calculated for filtering Mura regions. After image dilation, the proposed threshold is determined for detecting the non-uniform brightness region in TFT-LCD by means of monitoring the every pixel in the image. The experimental results obtained from several test images are used to illustrate the effectiveness and efficiency of the proposed method for Mura detection.

Development of a Dynamic System Simulating Pig Gastric Digestion

Chiang, C.-C.,Croom, J.,Chuang, S.-T.,Chiou, P.W.S.,Yu, B. Asian Australasian Association of Animal Productio 2008 Animal Bioscience Vol.21 No.10

The objective of this study was to develop a model for simulating gastric digestion in the pig. The model was constructed to include the chemical and physical changes associated with gastric digestion such as enzyme release, digestion product removal and gastric emptying. Digesta was collected from the stomach cannula of pigs to establish system parameters and to document the ability of the model to simulate gastric digestion. The results showed that the average pH of gastric digesta increased significantly from 2.47 to 4.97 after feed consumption and then decreased 140 min postprandial. The model described the decrease in pH within the pigs' stomach as $pH_t=5.182e^{-0.0014t}$, where t represents the postprandial time in minutes. The cumulative distribution function of liquid digesta was $V_t=64.509e^{0.0109t}$. The average pepsin activity in the liquid digesta was 317Anson units/mL. There was significant gastric emptying 220 min after feed consumption. The cybernetic dynamic system of gastric digestion was set according to the above data in order to compare with in vivo changes. The time course of crude protein digestion predicted by the model was highly correlated with observed in vivo digestion (r = 0.97; p = 0.0001), Model prediction for protein digestion was higher than that observed for a traditional static in vitro method (r = 0.89; p = 0.0001).

Nickel-Tin Solid-Liquid Inter-Diffusion Bonding

Chen-Chiang Yu,배성재,Pei-Chen Su,Tung-Han Chuang 한국정밀공학회 2014 International Journal of Precision Engineering and Vol.15 No.1

This paper presents a solid liquid inter-diffusion (SLID) bonding process using nickel-tin. By using two metals with different melting points, Ni3Sn4, an intermetallic compound (IMC) at 523K was formed. Unlike pure metal, the IMC can be used in high temperature applications of up to 800K without failure. The formation energy of Ni3Sn4 IMC was calculated to be 23.15 kJ/mol from the experimental result. Compared to the traditional soldering technique, a higher temperature resistance bonding joint can be achieved using the nickel-tin SLID bonding technique.

Optimization of Triple Response Systems by Using the Dual Response Approach and the Hooke-Jeeves Search Method

Fan, Shu-Kai S.,Huang, Chia-Fen,Chang, Ko-Wei,Chuang, Yu-Chiang Korean Institute of Industrial Engineers 2010 Industrial Engineeering & Management Systems Vol.9 No.1

This paper presents an extended computing procedure for the global optimization of the triple response system (TRS) where the response functions are nonconvex (nonconcave) quadratics and the input factors satisfy a radial region of interest. The TRS arising from response surface modeling can be approximated using a nonlinear mathematical program involving one primary (objective) function and two secondary (constraints) functions. An optimization algorithm named triple response surface algorithm (TRSALG) is proposed to determine the global optimum for the nondegenerate TRS. In TRSALG, the Lagrange multipliers of target (secondary) functions are computed by using the Hooke-Jeeves search method, and the Lagrange multiplier of the radial constraint is located by using the trust region (TR) method at the same time. To ensure global optimality that can be attained by TRSALG, included is the means for detecting the degenerate case. In the field of numerical optimization, as the family of TR approach always exhibits excellent mathematical properties during optimization steps, thus the proposed algorithm can guarantee the global optimal solution where the optimality conditions are satisfied for the nondegenerate TRS. The computing procedure is illustrated in terms of examples found in the quality literature where the comparison results with a gradient-based method are used to calibrate TRSALG.

Optimization of Triple Response Systems by Using the Dual Response Approach and the Hooke-Jeeves Search Method

Shu-Kai S. Fan,Chia-Fen Huang,Ko-Wei Chang,Yu-Chiang Chuang 대한산업공학회 2010 Industrial Engineeering & Management Systems Vol.9 No.1

This paper presents an extended computing procedure for the global optimization of the triple response system (TRS) where the response functions are nonconvex (nonconcave) quadratics and the input factors satisfy a radial region of interest. The TRS arising from response surface modeling can be approximated using a nonlinear mathematical program involving one primary (objective) function and two secondary (constraints) functions. An optimization algorithm named triple response surface algorithm (TRSALG) is proposed to determine the global optimum for the nondegenerate TRS. In TRSALG, the Lagrange multipliers of target (secondary) functions are computed by using the Hooke-Jeeves search method, and the Lagrange multiplier of the radial constraint is located by using the trust region (TR) method at the same time. To ensure global optimality that can be attained by TRSALG, included is the means for detecting the degenerate case. In the field of numerical optimization, as the family of TR approach always exhibits excellent mathematical properties during optimization steps, thus the proposed algorithm can guarantee the global optimal solution where the optimality conditions are satisfied for the nondegenerate TRS. The computing procedure is illustrated in terms of examples found in the quality literature where the comparison results with a gradient-based method are used to calibrate TRSALG.