Journal of the Korean Society for Precision Engineering

Real-time Instance Segmentation-based Object Detection and Adaptive Placing Algorithm for Low Cost Bin-picking System: Ki-Suk Kim, Hyun-Pyo Shin; J. Korean Soc. Precis. Eng. 2026;43(2):217-225.
Published online February 1, 2026; DOI: https://doi.org/10.7736/JKSPE.025.137

Abstract
PDF

Robots are increasingly utilized in manufacturing and logistics, where bin-picking has become crucial for managing randomly placed objects. However, traditional methods often rely on expensive 3D vision systems, have limited adaptability to unstructured environments, and primarily focus on the picking process, neglecting the placing tasks. To address these challenges, this study presents a cost-effective system that combines a depth camera, YOLO-based instance segmentation, and optimization-based inverse kinematics for real-time object detection and stable manipulation. In the placing stage, an adaptive algorithm detects empty tray holes and generates grid patterns, ensuring reliable placement even in the presence of tray misalignments, occupied slots, or partial occlusions. Experimental validation revealed a 91% success rate in mixed-object environments during picking tasks and a 94% success rate for placing tasks, even with tray displacement and occlusion conditions. The results demonstrate that the system maintains stable performance across both picking and placing processes while minimizing reliance on expensive hardware and complex initial setups. By enhancing flexibility and scalability, the proposed approach offers a practical solution for intelligent automation and can serve as a foundation for broader applications in assembly, logistics, and service robotics.

172 View
5 Download

Fabrication and Evaluation of CNT Spray Coated Strain Sensor: Yoon Ji Yum, Ji Hyun Park, Sang Hoon Lee; J. Korean Soc. Precis. Eng. 2026;43(2):197-206.
Published online February 1, 2026; DOI: https://doi.org/10.7736/JKSPE.025.116

Abstract
PDF

Carbon nanotubes (CNTs) are popular in strain sensors due to their exceptional electrical conductivity, flexibility, and sensitivity to deformation. In this study, a high-sensitivity strain sensor was fabricated by spray-coating CNT ink onto various paper substrates, with “lint-free paper” identified as the optimal choice. A total of 10 spray cycles ensured a reliable conductive coating. To enhance durability and broaden application potential, a PET protective layer was incorporated. The sensor's performance was assessed through bending tests using a push-pull gauge across a strain range of 0-2%. The lintfree paper-based sensor exhibited a consistent response up to 1.4% strain. The measured gauge factors (GF) were 121.370 in the 0-0.3% range, 70.999 in the 0.3-0.8% range, and 20.935 in the 0.8-1.4% range. A precise response was also noted when adjusting the bending angle in 1° increments, particularly within the 0-20° range. Additionally, the sensor was tested on the human wrist, confirming its viability for wearable applications. These findings indicate that the lint-free paper-based CNT strain sensor offers high sensitivity and measurement precision within narrow strain ranges. Its lightweight structure and flexible design suggest strong potential for practical use in areas such as sports monitoring and human motion detection.

152 View
3 Download

Ceiling Hazardous Object Inspection Robot for Counter-terrorism Security Check: Sangwoong Lee, Daegwon Koh, Meungsuk Lee, Hyeongseok Song, Juhyun Pyo, Jinho Suh, Murim Kim; J. Korean Soc. Precis. Eng. 2026;43(1):37-46.
Published online January 1, 2026; DOI: https://doi.org/10.7736/JKSPE.025.047

Abstract
PDF

Ceiling inspections present challenges due to limited accessibility and structural constraints. To ease the burden on security personnel, who would otherwise need to manually disassemble, inspect, and restore ceiling components, this study proposes a robotic system for detecting hazardous objects within ceiling environments. The proposed system features several key innovations: a hollow-structured track mechanism designed to reduce vibrations from jolting while traversing structural beams and to improve localization accuracy. We optimized the robot’s mass distribution and required drive torque through dynamic simulations to ensure stable mobility in confined ceiling spaces. For effective hazardous object detection, we developed a YOLOv8-Seg-based background learning algorithm that suppresses ceiling-structure patterns, allowing for the identification of unknown objects without prior class-specific training. Additionally, we introduced a frame-based filtering algorithm to enhance detection reliability by reducing false positives caused by motion blur during movement. The system's effectiveness was validated through experiments conducted in a ceiling-structured testbed, demonstrating its capability for accurate hazardous object detection under realistic operating conditions.

699 View
8 Download

A Study on the Detection of Hole in Automotive CV Joint Boot Using Image Processing and AI Techniques: Yun-Hyeok Lim, Hyeongill Lee; J. Korean Soc. Precis. Eng. 2025;42(10):861-869.
Published online October 1, 2025; DOI: https://doi.org/10.7736/JKSPE.025.050

Abstract
PDF

Detecting and analyzing defects in components or systems is crucial for maintaining high-quality standards in modern manufacturing and quality control. Recently, imaging-based defect detection methods have gained popularity across various engineering fields, highlighting their growing importance. Additionally, the integration of Artificial Intelligence (AI) to improve accuracy and efficiency is rapidly advancing. This paper presents a system that uses imaging to detect holes in CV joint boots, as these holes significantly affect the overall performance and durability of the system. Moreover, it introduces a method for enhancing detection performance by applying AI techniques. Validation tests on actual CV joint boots confirmed that the proposed method improves detection performance.

58 View
3 Download

A Highway Secondary Accident Prevention System based on FFT Analysis of Vehicle Collision Sounds: Minki Jung, Young Shin Cho, Yongsik Ham, Joong Bae Kim; J. Korean Soc. Precis. Eng. 2025;42(9):749-756.
Published online September 1, 2025; DOI: https://doi.org/10.7736/JKSPE.025.037

Abstract
PDF

This study introduces a highway secondary accident prevention system that employs Fast Fourier Transform (FFT) analysis of vehicle collision sounds. The system is designed to identify abnormal acoustic patterns produced during collisions and skidding events, enabling faster and more accurate accident detection than traditional methods. When a crash is detected, visual warning signals are instantly sent to nearby vehicles using LED devices powered by a photovoltaic panel and an energy storage system (ESS). Experimental results showed 100% detection accuracy during independent playback of collision, skidding, and driving sounds, and 80% accuracy during simultaneous playback. These results confirm the system's ability to effectively differentiate accident-related sounds and deliver timely alerts. This research offers an innovative and environmentally sustainable approach to enhancing highway safety and reducing the societal and economic consequences of secondary accidents.

61 View
3 Download

A Study on Polymer-based Cylindrical Flow Sensor for 2-dimensional Detection: Wonjun Lee, Sang Hoon Lee; J. Korean Soc. Precis. Eng. 2025;42(6):447-454.
Published online June 1, 2025; DOI: https://doi.org/10.7736/JKSPE.025.030

Abstract
PDF

In this study, we fabricated and investigated the polymer-based cylindrical flow sensor for two-dimensional (2D) detection. The flow sensor was the drag force type flowmeter which was fabricated with ecoflex. It had CNT/PDMS as the piezoresistive material and a cylindrical shape to measure the 2D flow. It also had impact resistance and ease of fabrication due to its polymer-based sensor. At first, two piezoresistive parts were applied to evaluate detection properties. Forces from various direction were applied. Results showed its potential as a sensing device. Following this, the final flow sensor was fabricated with four piezoresistive parts and its sensitivity was measured in the air flow from 0 to 30 m/s. Resistance changes were measured while rotating the sensor. Outputs showed a form of sine waves. Data were repeatedly collected under various conditions. The direction and air flow rate were then determined. To check physical impact resistance, a sudden high air flow rate with 100m/s was applied to the sensor and a stable output was obtained. These results suggest that such ecoflex-based cylindrical flow sensor can be used as a 2D flow rate sensor.

59 View
5 Download

Large-area Inspection Method for Machined Micro Hole Dimension Measurement Using Deep Learning in Silicon Cathodes: Jonghyeok Chae, Dongkyu Lee, Seunghun Oh, Yoojeong No; J. Korean Soc. Precis. Eng. 2025;42(2):139-145.
Published online February 1, 2025; DOI: https://doi.org/10.7736/JKSPE.024.117

Abstract
PDF

In this study, we propose a deep learning-based method for large-area inspection aimed at the high-speed detection of micro hole diameters. Micro holes are detected and stored in large images using YOLOv8, an object detection model. A super-resolution technique utilizing ESRGAN, an adversarial neural network, is applied to images of small micro holes, enhancing them to high resolution before measuring their diameters through image processing. When comparing the diameters measured after 8x super-resolution with the results from existing inspection equipment, the average error rate is remarkably low at 0.504%. The time taken to measure an image of one micro hole is 0.470 seconds, which is ten times faster than previous inspection methods. These results can significantly contribute to high-speed measurement and quality improvement through deep learning.

Citations

Citations to this article as recorded by

A Review of Intelligent Machining Process in CNC Machine Tool Systems
Joo Sung Yoon, Il-ha Park, Dong Yoon Lee
International Journal of Precision Engineering and Manufacturing.2025; 26(9): 2243. CrossRef

89 View
3 Download
Crossref

Remote Detection Technique of Trace Leak Gas based on Frequency Modulation Absorption Spectroscopy: Jungjae Park, Jae Yong Lee, Jae Heun Woo, Jonghan Jin; J. Korean Soc. Precis. Eng. 2024;41(10):741-746.
Published online October 1, 2024; DOI: https://doi.org/10.7736/JKSPE.024.087

Abstract
PDF

The LIDAR principle is used in a variety of fields, including large-scale pipeline facility management, industrial disaster safety control, and atmospheric environmental monitoring, to employ the remote gas detection technique. In this study, we designed and implemented a remote detection method for N2O gas leaks using absorption spectroscopy based on frequency modulation of a Mid-IR quantum cascade laser (QCL) with a wavelength of 4.5 μm. We direct the frequency-modulated beam, locked to a single absorption line of N2O, to a leak hole on a target surface within a range of approximately 50 m. For area scanning around the leak point, we use a galvano scanner to deflect the probe beam. The back-scattered beam from the diffuse target surface is then collected by a Cassegrain telescope with a diameter of 300 mm and detected by an InSb photo-detector with high photon sensitivity. To process the detected signal, we utilize fundamental and second harmonic detection with a lock-in amplifier, resulting in a relative gas concentration expressed as the second harmonic signal normalized by the fundamental signal. Our test results demonstrate that this proposed method can detect gas leaks as small as 0.005 sccm at a distance of 50 m.

41 View
0 Download

Trends in High Reflectance Coating Technologies for Cavity Ring-down Spectroscopy for Gas Detection: Haeng Yun Jung, June Park; J. Korean Soc. Precis. Eng. 2024;41(10):747-752.
Published online October 1, 2024; DOI: https://doi.org/10.7736/JKSPE.024.078

Abstract
PDF

Cavity ring-down spectroscopy (CRDS) is an ultra-sensitive direct absorption technique that offers unique advantages compared to other spectroscopic techniques. It can measure cooperative enhanced absorption for weakly absorbing species at ultra-low concentrations. This is achieved by leveraging the concept of a stable optical cavity, which allows for an effective optical path length of several kilometers within a small physical sample length. One advantage of CRDS technology is that it is unaffected by fluctuations in the intensity of the light source. Another advantage is its applicability to the detection of atoms, molecules, and radicals in the atmosphere. Additionally, the equipment associated with this technology is compact and robust. This paper will first introduce the fundamental principles and setup of CRDS technology. It will then provide an overview of the characteristics of the fabrication equipment and the high reflectivity mirror coating process used in cavity ring-down spectroscopy.

51 View
1 Download

A Study on the Measurement of Ship Hull Paint Thickness Using Collaborative Robots and Depth Cameras: Jun Jae Lee, Hyo Seok Lee, Hak Yi; J. Korean Soc. Precis. Eng. 2024;41(9):707-711.
Published online September 1, 2024; DOI: https://doi.org/10.7736/JKSPE.024.032

Abstract
PDF

This study presents a method for inspecting ship block wall painting using a cooperative robot. The robot used in this study is a representative example of a human-collaborative robot system. The end-effector of the robot is equipped with a depth camera, designed in an eye-in style. The camera is used to measure and evaluate the thickness of the paint applied to the iron plate, simulating the conditions of ship block wall painting. To improve the accuracy of the recognition, an object detection algorithm with rapid computation and high accuracy was utilized. The algorithm was used to identify and outline the paint areas using the Canny edge algorithm. The proposed method successfully demonstrated the precision of paint area recognition by clearly identifying the center point and outline of the areas. Comparing the paint thickness measurements with laser distance measurements confirmed the effectiveness of the proposed method.

46 View
1 Download

Defect Detection in the Forging Process of Wheel Nut Products through Object Detection: Chang Dae Kim, Seung Wook Baek, Wan Jjin Chung, Chang Whan Lee; J. Korean Soc. Precis. Eng. 2024;41(4):279-286.
Published online April 1, 2024; DOI: https://doi.org/10.7736/JKSPE.023.147

Abstract
PDF

This study developed a defect-detecting system for automotive wheel nuts. We proposed an image processing method using OpenCV for efficient defect-detection of automotive wheel nuts. Image processing method focused on noise removal, ratio adjustment, binarization, polar coordinate system formation, and orthogonal coordinate system conversion. Through data collection, preprocessing, object detection model training, and testing, we established a system capable of accurately classifying defects and tracking their positions. There are four defect types. Types 1 and 2 defects are defects of products where the product is completely broken circumferentially. Types 3 and 4 defects are defects are small circumferential dents and scratches in the product. We utilized Faster R-CNN and YOLOv8 models to detect defect types. By employing effective preprocessing and post-processing steps, we enhanced the accuracy. In the case of Fast RCNN, AP values were 0.92, 0.93, 0.76, and 0.49 for types 1, 2, 3, and 4 defects, respectively. The mAP was 0.77. In the case of YOLOv8, AP values were 0.78, 0.96, 0.8, and 0.51 for types for types 1, 2, 3, and 4 defects, respectively. The mAP was 0.76. These results could contribute to defect detection and quality improvement in the automotive manufacturing sector.

Citations

Citations to this article as recorded by

Large-area Inspection Method for Machined Micro Hole Dimension Measurement Using Deep Learning in Silicon Cathodes
Jonghyeok Chae, Dongkyu Lee, Seunghun Oh, Yoojeong Noh
Journal of the Korean Society for Precision Engineering.2025; 42(2): 139. CrossRef

83 View
0 Download
Crossref

Development of On-site Analytical Device for Hydrogen Sulfide Using Colorimetric Paper Sensor: Gi-Ja Lee, Yoo-Ri Na, Jae-Chul Lee; J. Korean Soc. Precis. Eng. 2024;41(1):11-18.
Published online January 1, 2024; DOI: https://doi.org/10.7736/JKSPE.023.072

Abstract
PDF

In this study, a highly sensitive analysis device for hydrogen sulfide that could be used quickly and easily on site was developed using a colorimetric paper sensor. To optimize analysis conditions, tests were performed for each function. Performances of the method using laboratory equipment and tools and the method using the developed device for hydrogen sulfide analysis were compared. The trend line of changes in parameter b of the image acquired by the on-site analytical device for hydrogen sulfide was calculated as y = 0.517x - 0.141 with a coefficient of determination (R2) of 0.9874. It was comparable to the method performed at the laboratory level, showing an excellent linearity. Using the calculated trend line as a calibration curve, the detection limit and quantification limit were found to be 2.386 μM and 7.952 μM, respectively. A reproducibility test showed a relative standard deviation of 5.7%, indicating a low dispersion of results.

49 View
0 Download

Anomaly Detection in a Combined Driving System based on Unsupervised Learning: Kichang Park, Yongkwan Lee; J. Korean Soc. Precis. Eng. 2023;40(11):921-928.
Published online November 1, 2023; DOI: https://doi.org/10.7736/JKSPE.023.068

Abstract
PDF

Anomaly detection models using big data generated from facilities and equipment have been adopted for predictive maintenance in the manufacturing industry. When facility faults or defects occur, different patterns of abnormal data are shown owing to their component behaviors. By detecting these pattern changes, it is possible to determine whether a facility abnormality occurs. This study evaluated the anomaly detection results from a combined driving system consisting of three driving motors for about six months at a manufacturing site. The learning data with an autoencoder model for about a month at the beginning of vibration data collection and continuous monitoring of anomalies using reconstruction errors showed that a component defect occurred in one driving motor, and the reconstruction error increased progressively about three months earlier than a facility manager found the failure. In addition, the micro-electro-mechanical systems sensor showed high amplitude in the entire frequency domain when high reconstruction errors occurred. However, the integrated electronics piezoelectric sensor showed different patterns as high amplitude in a specific frequency domain. The results of this study will be helpful for detecting facility abnormalities in combined driving systems using vibration sensors.

47 View
2 Download

A Study on Defect Detection Model of Bone Plates Using Multiple Filter CNN of Parallel Structure: 이송연 , 허용정; J. Korean Soc. Precis. Eng. 2023;40(9):677-683.
Published online September 1, 2023; DOI: https://doi.org/10.7736/JKSPE.022.106

Abstract
PDF

Bone plates are a medical device used for fixing broken bones, which should not have a crack and hole defect. Defect detection is very important because bone plate defect is very dangerous. In this study, we proposed a defect detection model based on a parallel type convolution neural network for detecting bone plate crack and pore deformation. All size filters were different according to the defect shape. A convolution neural network detected pore defects. Another convolution neural network detected the crack. Two convolution neural networks simultaneously detected different defect types. The performance of the defect detection model was measured and used for the F1- score. We confirmed that performance of the defect detection model was 98.4%. We confirmed that the defect detection time was 0.21 seconds.

39 View
0 Download

Development of Vision System for Quality Inspection of Machined Holes of Automobile Mechanical Parts: Min Yong Han, Ki Hyun Kim, Hyo Young Kim, Kwang In Ko, Kyo Mun Ku, Dong Ju Ki, Jae Hong Shim; J. Korean Soc. Precis. Eng. 2023;40(6):499-506.
Published online June 1, 2023; DOI: https://doi.org/10.7736/JKSPE.022.141

Abstract
PDF

This paper presents a vision system for machined hole quality inspection of mechanical parts in an automobile. Automobile parts have various shapes and holes created by press punches. However, if the press punch pin is broken, a hole is not created on the mechanical parts. This problem causes serious part quality defects. To solve this problem, we proposed a vision system that could easily and cheaply inspect the quality of holes in automotive machining parts. A software development environment was created to build an economical vision inspection system. Images were gathered using the Near-real-time method to overcome the low frame-per-second of inexpensive Complementary Metal Oxide Semiconductor (CMOS) webcams. Status of the hole was determined using template matching and distance between holes as a feature. The hardware required for vision inspection was designed so that it could be directly applied to the automotive part manufacturing process. When the proposed vision inspection system was tested by installing it in an automobile parts factory for 3 months, the system showed an inspection accuracy of at least 97.9%. This demonstrates the effectiveness of the proposed method with accuracy and speed of hole defect inspection of machined parts.