• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.1
• /
• pp.26-34
• /
• 1996

One of the challenging tasks in visual inspection using CCD camera is to identify surface defects in an image with complex textured backgeound. In microscopic view, the surface of real objects shows regular or random textured patterns. In this paper, we present a visual inspection algorithm to extract abnormal surface defects in an image with textured background. The algorithm uses the space and frequency information at the same time by introducing the Dominant Frequency Map(DFM) which can describe the frequency characteristics of every small local region of an input image. We demonstrate the feasibility and effectiveness of the method through a series of real experiments for a 14" TV CRT mold. The method successfully identifies a variety of infinitesimal defects, whose size is larger than $50\mu\textrm{m}$, of the mold. The experimental results show that the DFM based method is less sensitive to the environmental changes, such as illumination and defocusing, than conventional vision techniques.ques.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.2
• /
• pp.670-675
• /
• 2015

Medical imaging is one of the most powerful tools for gaining information about internal organs and tissues. It is a challenging task to develop sophisticated image analysis methods in order to improve the accuracy of diagnosis. The objective of this paper is to develop a Computer Aided Diagnostics (CAD) scheme for Brain Tumour detection from Magnetic Resonance Image (MRI) using active contour models and to investigate with several approaches for improving CAD performances. The problem in clinical medicine is the automatic detection of brain Tumours with maximum accuracy and in less time. This work involves the following steps: i) Segmentation performed by Fuzzy Clustering with Level Set Method (FCMLSM) and performance is compared with snake models based on Balloon force and Gradient Vector Force (GVF), Distance Regularized Level Set Method (DRLSE). ii) Feature extraction done by Shape and Texture based features. iii) Brain Tumour detection performed by various tree classifiers. Based on investigation FCMLSM is well suited segmentation method and Random Forest is the most optimum classifier for this problem. This method gives accuracy of 97% and with minimum classification error. The time taken to detect Tumour is approximately 2 mins for an examination (30 slices).

• Journal of the Korean Ceramic Society
• /
• v.23 no.1
• /
• pp.44-50
• /
• 1986

Si-ZnO n-n heterojunction diodes were prespared by r.f diode sputtering of the sintered ZnO target on n-type Si single crystal wafers and their structures and electrical properties were studied. The films were grown orientedly with the c-axis of crystallites perpendicular to the substrate surface at low r.f. powder and grown to polycrystalline films with random orientation at high r. f. powder. The crystallite size increased with the increasing substrate temperture The oriented texture films only were used to prepare the photovoltaic diodes and these didoes showed the photovoltaic effect veing positive of the ZnO side for the photons in the wavelength range of 380-1450nm. The sign reversal of phootovoltage which is the property os isotype heterojunction was not observed because of the degeneration of the ZnO films. The diode showed the forward rectification when it was biased with the ZnO side positive. The current-voltage characteristics exhibited the thermal-current type relationship J∝exp(qV/nkT) with n=1.23 at the low forward bias voltage and the tunnelling-current type relationship J∝exp($\alpha$V) where $\alpha$ was constant independent of temperature at the high forward bias voltage. The crystallite size of ZnO films were influenced largely on the photovoltaic properties of diodes ; The diodes with the films of the larger crystallites showed the poor photovoltaic properties. This reason may be cosidered that the ZnO films with the large crystallites could not grow to the electrically continuous films because the thickness of films was so thin in this experiment.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2012.05a
• /
• pp.100.2-100.2
• /
• 2012

Surface texturing of crystalline silicon is carried out in alkaline solutions for anisotropic etching that leads to random pyramids of about $10{\mu}m$ in size. Recently textured pyramids size gradually reduced using new solution. In this paper, we investigated that texture pyramids size had an impact on emitter property and front electrode (Ag) contact. To make small (${\sim}3{\mu}m$) and large (${\sim}10{\mu}m$) pyramids size, texturing times control and one side texturing using a silicon nitride film were carried out. Then formation and quality of POCl3-diffused n+ emitter in furnace compare with small and large pyramids by using SEM images, simulation (SILVACO, Athena module) and emitter saturation current density (J0e). After metallization, Ag contact resistance was measured by transfer length method (TLM) pattern. And surface distributions of Ag crystallites were observed by SEM images. Also, performance of cell which is fabricated by screen-printed solar cells is compared by light I-V.

• Journal of the Korean Applied Science and Technology
• /
• v.19 no.2
• /
• pp.131-136
• /
• 2002

As model waterborne acrylic coatings, mono-dispersed poly(butyl acrylate-methyl methacrylate) copolymer latexes of random copolymer and core/shell type graft copolymer were prepared by seeded multi-staged emulsion polymerization with particle size of $180{\sim}200$ nm using semi-batch type process. Sodium lauryl sulfate and potassium persulfate were used as an emulsifier and an initiator, respectively. The effect of particle texture including core/shell phase ratio, glass transition temperature and crosslinking density, and film forming temperature on the film formation and final properties of film was investigated using SEM, AFM, and UV in this study. The film formation behavior of model latex was traced simultaneously by the weight loss measurement and by the change of tensile properties and UV transmittance during the entire course of film formation. It was found that the increased glass transition temperature and higher crosslinking degree of latex resulted in the delay of the onset of coalescence of particles by interdiffusion during film forming process. This can be explained qualitatively in terms of diffusion rate of polymer chains. However, the change of weight loss during film formation was insensitive to discern each film forming stages-I, II and III.

• International Journal of Computer Science & Network Security
• /
• v.21 no.1
• /
• pp.40-48
• /
• 2021

Content-Based Image Retrieval (CBIR) system plays a vital role to retrieve the relevant images as per the user perception from the huge database is a challenging task. Images are represented is to employ a combination of low-level features as per their visual content to form a feature vector. To reduce the search time of a large database while retrieving images, a novel image retrieval technique based on feature dimensionality reduction is being proposed with the exploit of metaheuristic optimization techniques based on Genetic Algorithm (GA), Extended Binary Cuckoo Search (EBCS) and Whale Optimization Algorithm (WOA). Each image in the database is indexed using a feature vector comprising of fuzzified based color histogram descriptor for color and Median binary pattern were derived in the color space from HSI for texture feature variants respectively. Finally, results are being compared in terms of Precision, Recall, F-measure, Accuracy, and error rate with benchmark classification algorithms (Linear discriminant analysis, CatBoost, Extra Trees, Random Forest, Naive Bayes, light gradient boosting, Extreme gradient boosting, k-NN, and Ridge) to validate the efficiency of the proposed approach. Finally, a ranking of the techniques using TOPSIS has been considered choosing the best feature selection technique based on different model parameters.

• Journal of Soil and Groundwater Environment
• /
• v.21 no.6
• /
• pp.179-191
• /
• 2016

In addition to the measurement of the concentration of soil contaminants, the new idea of indicative parameters was proposed to validate the remedial works through the monitoring for the changes of soil characteristics after applying the clean up technologies. The parameters like CFU (colony forming unit), pH and soil texture were recommended as indicative parameters for land farming. In case of soil washing, water content and the particle size distribution of the sludge were recommended as indicative parameters. The sludge is produced through the particle separation process in soil washing and it is usually treated as a waste. The parameters like water content, organic matter content, CEC (cation exchange capacity) and CFU were recommended as indicative parameters for the low temperature thermal desorption method. Besides the indicative parameter, sampling methods in stock pile and the optimal minimum amount of composite soil sample were proposed. The rates of sampling error in regular grid, zigzag, four bearing, random grid methods were 17.3%, 17.6%, 17.2% and 16.5% respectively. The random grid method showed the minimum sampling error among the 4 kinds of sampling methods although the differences in sampling errors were very little. Therefore the random grid method was recommended as an appropriate sampling method in stock pile. It was not possible to propose a value of optimal minimum amount of composite soil sample based on the real analytical data due to the dynamic variation of $CV_{fund{\cdot}error}$. Instead of this, 355 g of soil was recommended for the optimal minimum amount of composite soil sample under the assumption of ISO 10381-8.

• Korean Journal of Materials Research
• /
• v.4 no.8
• /
• pp.895-905
• /
• 1994

The solid phase crystallization behavior of undoped amorphous $Si_{1-x}Ge_{x}$ (X=O to 0.53) alloyfilms was studied by X-ray diffractometry(XRD) and transmission electron microscopy(TEM). Thefilms were deposited on thermally oxidized 5" (100) Si wafer by MBE(Mo1ecular Beam Epitaxy) at 300'C and annealed in the temperature range of $500^{\circ}C$ ~ $625^{\circ}C$. From XRD results, it was found that the thermal budget for full crystallization of the film is significantly reduced as the Ge concentration in thefilm is increased. In addition, the results also shows that pure amorphous Si film crystallizes with astrong (111) texture while the $Si_{1-x}Ge_{x}$ alloy film crystallzes with a (311) texture suggesting that the solidphase crystallization mechanism is changed by the incorporation of Ge. TEM analysis of the crystallized filmshow that the grain morphology of the pure Si is an elliptical and/or a dendrite shape with high density ofcrystalline defects in the grains while that of the $Si_{0.47}Ge_{0.53}$ alloy is more or less equiaxed shape with muchlower density of defects. From these results, we conclude that the crystallization mechanism changes fromtwin-assisted growth mode to random growth mode as the Ge cocentration is increased.ocentration is increased.

• Journal of Life Science
• /
• v.21 no.9
• /
• pp.1251-1258
• /
• 2011

Multiparous Hanwoo cull cows have been generally regarded to have poor meat quality compared to young and primiparous Hanwoo in Korea, and there have been few studies attempting to understand various feeding programs for the multiparous Hanwoo cull cow. In this study, the effects of a feeding program consisting of two different diets for multiparous Hanwoo cull cows on meat production and quality were tested in comparison to a commercially used diet. Diets for treatment consisted of two levels of crude protein contents, 14.28% and 12.70% for early fattening and finishing, respectively. For a control, commercially used fattening feed (12.39% crude protein) was used. Feeding trials were performed at three different farms. In farm A, 29 herds of multiparous Hanwoo cull cows were used for the treatment group and 3 herds for the control. In farms B and C, the number of animal herds for treatment and control were 8 vs. 3 and 11 vs. 4 herds, respectively. Experiment diets were fed for an average of 211 days and in treatment, early fattening diet was fed for 4 months and then finishing feed was fed until slaughter. Average daily gains, thickness of back fat, area of Longissimus dorsi, carcass weight, index of carcass weight, intramuscular fat, meat color and texture were analyzed after slaughter. Random effect model [8] was employed in effect analysis. Positive effects of treatment were found in terms of average daily gain, back fat thickness, Longissimus dorsi area, carcass weight, and intramuscular fat. Carcass yield index, meat color and texture showed a negative effect. In this study, significant results were not found in all factors analyzed because the variance between experiment farms was large, however the 90% confidence interval of summary effects of ADG, back fat thickness, Longissimus dorsi area and carcass weight were significant and that of carcass yield index, intramuscular fat, meat color and texture were less so. Conclusively, a high protein diet fed early during the fattening period in multiparous Hanwoo cull cows could have positive effects on meat production.

• Journal of Internet Computing and Services
• /
• v.21 no.6
• /
• pp.23-31
• /
• 2020

This study was carried out to generate various images of railroad surfaces with random defects as training data to be better at the detection of defects. Defects on the surface of railroads are caused by various factors such as friction between track binding devices and adjacent tracks and can cause accidents such as broken rails, so railroad maintenance for defects is necessary. Therefore, various researches on defect detection and inspection using image processing or machine learning on railway surface images have been conducted to automate railroad inspection and to reduce railroad maintenance costs. In general, the performance of the image processing analysis method and machine learning technology is affected by the quantity and quality of data. For this reason, some researches require specific devices or vehicles to acquire images of the track surface at regular intervals to obtain a database of various railway surface images. On the contrary, in this study, in order to reduce and improve the operating cost of image acquisition, we constructed the 'Defective Railroad Surface Regeneration Model' by applying the methods presented in the related studies of the Generative Adversarial Network (GAN). Thus, we aimed to detect defects on railroad surface even without a dedicated database. This constructed model is designed to learn to generate the railroad surface combining the different railroad surface textures and the original surface, considering the ground truth of the railroad defects. The generated images of the railroad surface were used as training data in defect detection network, which is based on Fully Convolutional Network (FCN). To validate its performance, we clustered and divided the railroad data into three subsets, one subset as original railroad texture images and the remaining two subsets as another railroad surface texture images. In the first experiment, we used only original texture images for training sets in the defect detection model. And in the second experiment, we trained the generated images that were generated by combining the original images with a few railroad textures of the other images. Each defect detection model was evaluated in terms of 'intersection of union(IoU)' and F1-score measures with ground truths. As a result, the scores increased by about 10~15% when the generated images were used, compared to the case that only the original images were used. This proves that it is possible to detect defects by using the existing data and a few different texture images, even for the railroad surface images in which dedicated training database is not constructed.