• Journal of Biosystems Engineering
• /
• 제26권6호
• /
• pp.571-578
• /
• 2001

Production of green pepper has increased for ten years in Korea, as customer's preference of a pepper tuned to fiesta one. This study was conducted to develop an on-line fading algorithm of green pepper using machine vision and aimed to develop the automatic on-line grading and sorting system. The machine vision system was composed of a professive scan R7B CCD camera, a frame grabber and sets of 3-wave fluorescent lamps. The length and curvature, which were main quality factors of a green pepper were measured while removing the stem region. The first derivative of the thickness profile was used to remove the stem area of the segmented image of the pepper. A new boundary was generated after the stem was removed and a baseline of a pepper which was used for the curvature determination was also generated. The developed algorithm showed that the accuracy of the size measurement was 86.6% and the accuracy of the bent was 91.9%. Processing time spent far grading was around 0.17 sec per pepper.

• 융합신호처리학회논문지
• /
• 제21권4호
• /
• pp.154-161
• /
• 2020

본 논문에서는 인간의 노력을 줄이고 정확성을 높이기 위해 사과의 색을 기반으로 하는 분류 시스템을 제안하였다. 제안된 분류 시스템은 카메라, 모터 및 라즈베리 파이로 구성되어 있고, 미성숙, 성숙, 익은 등으로 총 4가지 종류의 사과를 분류할 수 있다. 시장에서 다양한 종류의 사과를 100개 구입하여 무작위로 선택하여 평가하였다. 정확도는 95%였고 처리 시간은 사과당 약 8초였다. 제안한 시스템은 인력 감축에 유용할 것으로 예상된다.

• Journal of the Korean Wood Science and Technology
• /
• 제39권2호
• /
• pp.163-171
• /
• 2011

Because Japanese cedar shows lower mechanical performance, glued-laminated timber (glulam) can be a better way to utilize Japanese cedar for structural purpose. However, low yield of higher grade lamination from log makes it difficult to design structural glulam. This study was aimed to increase the yield of higher grade lamination and provide higher efficiency of manufacturing structural lamination by ultrasonic log sorting technology. Logs were sorted by an existing log grading rule regulated by Korea Forest Research Institute (KFRI). It was found that the KFRI log grading rule contributed to finding better logs in viewpoint of the volumetric yield and it can reduce the number of rejected lumber by visual grading. However, it could not identify better logs to produce higher-grade products. To find an appropriate log-sorting-method for structural products, log diameter and ultrasonic time of flight (TOF) for the log were considered as factors to affect mechanical performance of resulting products. However, it was found that influence of log diameter on mechanical performance of resulting products was very small. The TOF showed a possibility to sort logs by mechanical performance of resulting products even though a coefficient of correlation was not strong (R = 0.6). In a case study, the log selection based on the ultrasonic TOF of the log increased the yield of the outermost tension lamination (E8 or better grade, KS F 3021) from 2.6% to 12.5% and reduced LTE5 (lower than E5 grade) lamination from 43.6% to 10.3%, compared with the existing KFRI log grading rule.

• 자원리싸이클링
• /
• 제30권4호
• /
• pp.3-10
• /
• 2021

도시광산의 대표적인 사례인 폐 전자부품 재활용 분야는 유용한 선별기술의 개발이 필요하다. 본 연구에서는 전자부품들을 4개의 그룹으로 선별하기 위해 딥러닝에 의한 이미지 식별을 기반으로 한 선별기를 개발하였다. 전자부품들은 폐 인쇄회로기판으로부터 회수되었으며, 처리 후 차이점에 따라 분리되어야 한다. 선별기는 GPU, 카메라, 컨베이어벨트, 공기 압축기로 이루어진 워크스테이션으로 구성되어있다. 컨베이어벨트(벨트 속도: 6cm/s) 위에 있는 작은 크기(3.5cm 이하)의 전자부품들을 데이터로 가져와 학습했다. 이미지 식별의 정확도는 종류에 따라 96%, 그리고 그룹에 따라 99%였다. 벨트속도와 압축공기의 블로다운 시간과 같은 작동조건에 변화를 주었을 때, 이미지 식별의 정확도 그리고 블로다운에 의한 회수율을 측정하여 선별의 최적 조건을 결정했다. 최적조건에서 이미지 분류의 정확도는 그룹별로 98.7%였다. 선별 효율은 70% 이상이었다.

• 한국군사과학기술학회지
• /
• 제26권5호
• /
• pp.384-391
• /
• 2023

In modern electronic warfare, a number of radar emitters are in operation, causing radar receivers to receive high-density signal pulses that occur simultaneously. To analyze the radar signals more accurately and identify enemies, the sorting process of high-density radar signals is very important before analysis. Recently, machine learning algorithms, specifically K-means clustering, are the subject of research aimed at improving the accuracy of radar signal sorting. One of the challenges faced by these studies is that the clustering results can vary depending on how the initial points are selected and how many clusters number are set. This paper introduces a repeated K-means clustering algorithm that aims to accurately cluster all data by identifying and addressing false clusters in the radar sorting problem. To verify the performance of the proposed algorithm, experiments are conducted by applying it to simulated signals that are generated by a signal generator.

• 한국자원리싸이클링학회:학술대회논문집
• /
• 한국자원리싸이클링학회 2005년도 춘계임시총회 및 제25회 학술발표대회
• /
• pp.57-61
• /
• 2005

본 연구에서는 색도인식이 가능한 머신 비젼시스템을 설계 제작하고 제작한 시스템을 이용하여 철스크랩에 혼합되어 있는 Cu 스크랩을 자동으로 분리하는 시험연구를 행하였다. 머신비젼시스템은 크게 측정부, 이송부 및 ejector로 구분되며 CCD카메라, 광원 및 frame grabber로 구성된 측정부에서 이메지 프로쎄싱 알고리즘에 의해 이송되어 오는 스크랩 표면의 색도를 인식하게 된다. 또한 그 인식결과에 따라 I/O interface가 구성된 콘트롤러에 의해 에어노즐을 작동하도록 하여 임의로 지정한 특정한 표면색상의 스크랩만을 분리하도록 되어 있다. 본 시스템을 이용하여 철스크랩에 포함된 Cu스크랩의 분리시험을 행한 결과, 스크랩 이송속도가 15m/min.에서 90% 이상의 분리효율을 나타내어 향후 분리효율의 고속화가 실현될 경우 산업적으로 적용가능성이 매우 높은 것으로 판단되었다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2005년도 춘계학술대회 논문집
• /
• pp.997-1000
• /
• 2005

Dynamic job shop environment requires not only more flexible capabilities of a CAPP system but higher utility of the generated process plans. In order to meet the requirements, this paper develops an algorithm that can select machines for the machining operations to be performed by predicting the machine loads. The developed algorithm is based on the multiple objective genetic algorithm that gives rise to a set of optimal solutions (in general, known as Pareto-optimal solutions). The objective shows a combination of the minimization of part movement and the maximization of machine utility balance. The algorithm is characterized by a new and efficient method for nondominated sorting, which can speed up the running time, as well as a method of two stages for genetic operations, which can maintain a diverse set of solutions. The performance of the algorithm is evaluated by comparing with another multiple objective genetic algorithm, called NSGA-II.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2005년도 BIOINFO 2005
• /
• pp.256-259
• /
• 2005

Among non-synonymous SNPs that cause amino acid change in the protein product, the selection of disease-causing SNPs has been of great interest. We present the comparison between the evolutionary (SIFT score) and structural information (binding pocket) to show that the incorporation between them provides an advantage of sorting disease-causing SNPs from normal SNPs. To set up the procedure, we apply the machine learning method to the test data set from the laboratory experiments.

• 한국농업기계학회:학술대회논문집
• /
• 한국농업기계학회 2001년도 하계 학술대회 논문집
• /
• pp.308-313
• /
• 2001

• 한국정밀공학회지
• /
• 제24권4호
• /
• pp.84-92
• /
• 2007

Dynamic process planning requires not only more flexible capabilities of a CAPP system but also higher utility of the generated process plans. In order to meet the requirements, this paper develops an algorithm that can select machines for the machining operations by calculating the machine loads. The developed algorithm is based on the multi-objective genetic algorithm that gives rise to a set of optimal solutions (in general, known as the Pareto-optimal solutions). The objective is to satisfy both the minimization number of part movements and the maximization of machine utilization. The algorithm is characterized by a new and efficient method for nondominated sorting through K-means algorithm, which can speed up the running time, as well as a method of two stages for genetic operations, which can maintain a diverse set of solutions. The performance of the algorithm is evaluated by comparing with another multiple objective genetic algorithm, called NSGA-II and branch and bound algorithm.