• Korean Journal of Heritage: History & Science
• /
• v.44 no.4
• /
• pp.142-153
• /
• 2011

'Silla Monument Stone in Jungseong-ri, Pohang' was discovered in Pohang City, Gyeongsangbuk Province of Korea in 2009. The monument stone with irregular shape has dimensions of maximum height of 105cm, width of 47.6~49.6cm, thickness of 13.8~14.7cm and weight of 115kg. The results of monument stone was found to be granitite. Conservation treatment procedure was carried out in the order of production of Weathering map, cleaning of surface pollutants, consolidation using ethyl silicate. Black pollutant(crust) that covered more than 60% of the surface was analyzed first in order to remove the pollutants on the surface of the monumental stone by cleaning of surface pollutants using laser. The purpose on analysis was not only to verify the pollutants on the stone but also to carry out preliminary cleaning by securing rocks with same pollutant as the monumental stone. As the results of analysis using p-XRF(PMI. INNOV-X, USA) on the site, high level of Mn and Fe were detected, and the analysis of small section that had been fallen off with SEM/EDX for the purpose of quantitative analysis also detected high level of Mn. The Similar contaminants on Stone secured in the manner described above were made into 10 samples ($5cm{\times}5cm$) and was subjected to preliminary cleaning by Nd-YAG Laser(Laser&Physics, Korea). The results of surface observation through portable microscope during cleaning revealed that the power of 460mJ, wavelength of 1064nm and irradiation frequency of 1,800~2,300 per $25cm^2$ were most effective. Evaluation on the preservative treatment was made through confirmation of the extent of removal through color-difference meter measurement and component analysis prior to and following removal of the pollutants. As the result of color-difference meter measurement increase in the brightness was evidenced by the increase in the brightness ($L^*$)value from 33 to 49, and it was possible to ascertain the reduction in the pollutants as the content of Mn was reduced by about 80% from $50,000{\pm}5,000ppm$ to $10,000{\pm}2,000pmm$ from the result of component analysis.

• Journal of Bio-Environment Control
• /
• v.32 no.2
• /
• pp.106-114
• /
• 2023

With rising concerns about pesticide spray drift by aerial application, this study attempt to evaluate aerodynamic property and collection efficiency of spray drift according to the leaf area index (LAI) of crop for preventing undesirable pesticide contamination by the spray-drift tunnel experiment. The collection efficiency of the plant with 'Low' LAI was measured at 16.13% at a wind speed of 1 m·s^-1. As the wind speed increased to 2 m·s^-1, the collection efficiency of plant with the same LAI level increased 1.80 times higher to 29.06%. For the 'Medium' level LAI, the collection efficiency was 24.42% and 43.06% at wind speed of 1 m·s^-1 and 2 m·s^-1, respectively. For the 'High' level LAI, it also increased 1.24 times higher as the wind speed increased. The measured results indicated that the collection of spray droplets by leaves were increased with LAI and wind speed. This also implied that dense leaves would have more advantages for preventing the drift of airborne spray droplets. Aerodynamic properties also tended to increase as the LAI increased, and the regression analysis of quadric equation and power law equation showed high explanatory of 0.96-0.99.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.