Limits...
Non-destructive inspection methods for LEDs using real-time displaying Optical Coherence Tomography.

Cho NH, Jung U, Kim S, Kim J - Sensors (Basel) (2012)

Bottom Line: The SD-OCT and SS-OCT images were compared with each other in the same sample to study their advantages.In addition, the volume of the fluorophore space was calculated from the OCT images.We expect this method can improve the inspection efficacy over traditional inspection methods such as Charged Coupled Device (CCD) camera or X-ray instruments.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical Engineering and Computer Science, Kyungpook National University, 1370, Sankyuk-dong, Buk-gu, Daegu 702-701, Korea. nhcho@knu.ac.kr

ABSTRACT
In this study, we report the applicability of two different Optical Coherence Tomography (OCT) technologies for inspecting Light Emitting Diode (LED) structures. Sectional images of a LED were captured using a Spectral Domain OCT (SD-OCT) system and a Swept Source OCT (SS-OCT) system. Their center wavelengths are 850 and 1,310 nm, respectively. We acquired cross-sectional two dimensional (2D) images of a normal LED and extracted sectional profiles to inspect possible wire disconnection that may be present in the LED manufacturing process. The SD-OCT and SS-OCT images were compared with each other in the same sample to study their advantages. The distribution of fluorescence material was observed more clearly with the SD-OCT of 850 nm wavelength, whereas the status of wire connection was clearer in the SS-OCT images with 1,310 nm wavelength. In addition, the volume of the fluorophore space was calculated from the OCT images. This is the first report that a nondestructive optical imaging modality such as OCT can be applied to finding screen defects in LED. We expect this method can improve the inspection efficacy over traditional inspection methods such as Charged Coupled Device (CCD) camera or X-ray instruments.

No MeSH data available.


Related in: MedlinePlus

Program block diagram. (a) SD-OCT Program block diagram; (b) SS-OCT Program block diagram.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3472834&req=5

f4-sensors-12-10395: Program block diagram. (a) SD-OCT Program block diagram; (b) SS-OCT Program block diagram.

Mentions: The driving software for both the SD and SS-OCT systems were programed using Compute Unified Device Architecture (CUDA) version 3.2 for Visual Studio 2008 and Graphics Processing Unit (GPU) programming. Figure 4(a) is the block diagram of the SD-OCT program. To improve the data acquisition speed, we used a personal computer with two main memories (RAM) as buffers. Figure 4(b) is the structure diagram that shows the procedure to display images from the memory assigned to the digitizer. During this procedure, the data are acquired from the digitizer and copied to the RAM. We adapted the GPU to provide fast image processing and displaying, which are important issues in the fault inspection environment. With GPU processing, we could realize the real-time display feature after performing massive data processing, including k-domain linearization, background removal, Fast Fourier Transform (FFT), and log scaling processes [21–23].


Non-destructive inspection methods for LEDs using real-time displaying Optical Coherence Tomography.

Cho NH, Jung U, Kim S, Kim J - Sensors (Basel) (2012)

Program block diagram. (a) SD-OCT Program block diagram; (b) SS-OCT Program block diagram.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3472834&req=5

f4-sensors-12-10395: Program block diagram. (a) SD-OCT Program block diagram; (b) SS-OCT Program block diagram.
Mentions: The driving software for both the SD and SS-OCT systems were programed using Compute Unified Device Architecture (CUDA) version 3.2 for Visual Studio 2008 and Graphics Processing Unit (GPU) programming. Figure 4(a) is the block diagram of the SD-OCT program. To improve the data acquisition speed, we used a personal computer with two main memories (RAM) as buffers. Figure 4(b) is the structure diagram that shows the procedure to display images from the memory assigned to the digitizer. During this procedure, the data are acquired from the digitizer and copied to the RAM. We adapted the GPU to provide fast image processing and displaying, which are important issues in the fault inspection environment. With GPU processing, we could realize the real-time display feature after performing massive data processing, including k-domain linearization, background removal, Fast Fourier Transform (FFT), and log scaling processes [21–23].

Bottom Line: The SD-OCT and SS-OCT images were compared with each other in the same sample to study their advantages.In addition, the volume of the fluorophore space was calculated from the OCT images.We expect this method can improve the inspection efficacy over traditional inspection methods such as Charged Coupled Device (CCD) camera or X-ray instruments.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical Engineering and Computer Science, Kyungpook National University, 1370, Sankyuk-dong, Buk-gu, Daegu 702-701, Korea. nhcho@knu.ac.kr

ABSTRACT
In this study, we report the applicability of two different Optical Coherence Tomography (OCT) technologies for inspecting Light Emitting Diode (LED) structures. Sectional images of a LED were captured using a Spectral Domain OCT (SD-OCT) system and a Swept Source OCT (SS-OCT) system. Their center wavelengths are 850 and 1,310 nm, respectively. We acquired cross-sectional two dimensional (2D) images of a normal LED and extracted sectional profiles to inspect possible wire disconnection that may be present in the LED manufacturing process. The SD-OCT and SS-OCT images were compared with each other in the same sample to study their advantages. The distribution of fluorescence material was observed more clearly with the SD-OCT of 850 nm wavelength, whereas the status of wire connection was clearer in the SS-OCT images with 1,310 nm wavelength. In addition, the volume of the fluorophore space was calculated from the OCT images. This is the first report that a nondestructive optical imaging modality such as OCT can be applied to finding screen defects in LED. We expect this method can improve the inspection efficacy over traditional inspection methods such as Charged Coupled Device (CCD) camera or X-ray instruments.

No MeSH data available.


Related in: MedlinePlus