• Korean Journal of Remote Sensing
• /
• v.39 no.5_1
• /
• pp.669-681
• /
• 2023

Studies have tried to apply remote sensing technology, a non-destructive survey method, instead of the existing destructive survey, which requires relatively large labor input and a long time to estimate chlorophyll content, which is an important indicator for evaluating the growth of fruit trees. This study was conducted to non-destructively evaluate the chlorophyll content of pear tree leaves using unmanned aerial vehicle-based hyperspectral imagery for two years(2021, 2022). The reflectance of the single bands of the pear tree canopy extracted through image processing was band rationed to minimize unstable radiation effects depending on time changes. The estimation (calibration and validation) models were developed using machine learning algorithms of elastic-net, k-nearest neighbors(KNN), and support vector machine with band ratios as input variables. By comparing the performance of estimation models based on full band ratios, key band ratios that are advantageous for reducing computational costs and improving reproducibility were selected. As a result, for all machine learning models, when calibration of coefficient of determination (R²)≥0.67, root mean squared error (RMSE)≤1.22 ㎍/cm², relative error (RE)≤17.9% and validation of R²≥0.56, RMSE≤1.41 ㎍/cm², RE≤20.7% using full band ratios were compared, four key band ratios were selected. There was relatively no significant difference in validation performance between machine learning models. Therefore, the KNN model with the highest calibration performance was used as the standard, and its key band ratios were 710/714, 718/722, 754/758, and 758/762 nm. The performance of calibration showed R²=0.80, RMSE=0.94 ㎍/cm², RE=13.9%, and validation showed R²=0.57, RMSE=1.40 ㎍/cm², RE=20.5%. Although the performance results based on validation were not sufficient to estimate the chlorophyll content of pear tree leaves, it is meaningful that key band ratios were selected as a standard for future research. To improve estimation performance, it is necessary to continuously secure additional datasets and improve the estimation model by reproducing it in actual orchards. In future research, it is necessary to continuously secure additional datasets to improve estimation performance, verify the reliability of the selected key band ratios, and upgrade the estimation model to be reproducible in actual orchards.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_1
• /
• pp.1413-1425
• /
• 2023

The increasing frequency of wildfires due to climate change is causing extreme loss of life and property. They cause loss of vegetation and affect ecosystem changes depending on their intensity and occurrence. Ecosystem changes, in turn, affect wildfire occurrence, causing secondary damage. Thus, accurate estimation of the areas affected by wildfires is fundamental. Satellite remote sensing is used for forest fire detection because it can rapidly acquire topographic and meteorological information about the affected area after forest fires. In addition, deep learning algorithms such as convolutional neural networks (CNN) and transformer models show high performance for more accurate monitoring of fire-burnt regions. To date, the application of deep learning models has been limited, and there is a scarcity of reports providing quantitative performance evaluations for practical field utilization. Hence, this study emphasizes a comparative analysis, exploring performance enhancements achieved through both model selection and data design. This study examined deep learning models for detecting wildfire-damaged areas using Landsat 8 satellite images in California. Also, we conducted a comprehensive comparison and analysis of the detection performance of multiple models, such as U-Net and High-Resolution Network-Object Contextual Representation (HRNet-OCR). Wildfire-related spectral indices such as normalized difference vegetation index (NDVI) and normalized burn ratio (NBR) were used as input channels for the deep learning models to reflect the degree of vegetation cover and surface moisture content. As a result, the mean intersection over union (mIoU) was 0.831 for U-Net and 0.848 for HRNet-OCR, showing high segmentation performance. The inclusion of spectral indices alongside the base wavelength bands resulted in increased metric values for all combinations, affirming that the augmentation of input data with spectral indices contributes to the refinement of pixels. This study can be applied to other satellite images to build a recovery strategy for fire-burnt areas.

• Journal of Korea Water Resources Association
• /
• v.57 no.1
• /
• pp.35-44
• /
• 2024

Research in dam inflow prediction has actively explored the utilization of data-driven machine learning and deep learning (ML&DL) tools across diverse domains. Enhancing not just the inherent model performance but also accounting for model characteristics and preprocessing data are crucial elements for precise dam inflow prediction. Particularly, existing rainfall data, derived from snowfall amounts through heating facilities, introduces distortions in the correlation between snow accumulation and rainfall, especially in dam basins influenced by snow accumulation, such as Soyang Dam. This study focuses on the preprocessing of rainfall data essential for the application of ML&DL models in predicting dam inflow in basins affected by snow accumulation. This is vital to address phenomena like reduced outflow during winter due to low snowfall and increased outflow during spring despite minimal or no rain, both of which are physical occurrences. Three machine learning models (SVM, RF, LGBM) and two deep learning models (LSTM, TCN) were built by combining rainfall and inflow series. With optimal hyperparameter tuning, the appropriate model was selected, resulting in a high level of predictive performance with NSE ranging from 0.842 to 0.894. Moreover, to generate rainfall correction data considering snow accumulation, a simulated snow accumulation algorithm was developed. Applying this correction to machine learning and deep learning models yielded NSE values ranging from 0.841 to 0.896, indicating a similarly high level of predictive performance compared to the pre-snow accumulation application. Notably, during the snow accumulation period, adjusting rainfall during the training phase was observed to lead to a more accurate simulation of observed inflow when predicted. This underscores the importance of thoughtful data preprocessing, taking into account physical factors such as snowfall and snowmelt, in constructing data models.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.16 no.2
• /
• pp.110-114
• /
• 2012

Purpose : PET/CT is widely used for early checking up of cancer and following up of pre and post operation. Image reconstruction method is advanced with mechanical function. We want to evaluate image quality of each reconstruction program based on time of flight (TOF). Materials and Methods : After acquiring phantom images during 2 minutes with Gemini TF (Philips, USA), Biograph mCT (Siemens, USA) and Discovery 690 (GE, USA), we reconstructed image applied to Astonish TF (Philips, USA), ultraHD PET (Siemens, USA), Sharp IR (GE, USA) and not applied. inside of Flangeless Esser PET phantom (Data Spectrum corp., USA) was filled with $^{18}F$-FDG 1.11 kBq/ml (30 Ci/ml) and 4 hot inserts (8. 12. 16. 25 mm) were filled with 8.88 kBq/ml (240 ${\mu}Ci/ml$) the ratio of background activity and hot inserts activity was 1 : 8. Inside of triple line phantom (Data Spectrum corp., USA) was filled with $^{18}F$-FDG 37 MBq/ml (1 mCi). Three of lines were filled with 0.37 MBq (100 ${\mu}Ci$). Contrast ratio and background variability were acquired from reconstruction image used Flangeless Esser PET phantom and resolution was acquired from reconstruction image used triple line phantom. Results : The contrast ratio of image which was not applied to Astonish TF was 8.69, 12.28, 19.31, 25.80% in phantom lid of which size was 8, 12, 16, 25 mm and it which was applied to Astonish TF was 6.24, 13.24, 19.55, 27.60%. It which was not applied to ultraHD PET was 4.94, 12.68, 22.09, 30.14%, it which was applied to ultraHD PET was 4.76, 13.23, 23.72, 31.65%. It which was not applied to SharpIR was 13.18, 17.44, 28.76, 34.67%, it which was applied to SharpIR was 13.15, 18.32, 30.33, 35.73%. The background variability of image which was not applied to Astonish TF was 5.51, 5.42, 7.13, 6.28%. it which was applied to Astonish TF was 7.81, 7.94, 6.40 6.28%. It which was not applied to ultraHD PET was 6.46, 6.63, 5.33, 5.21%, it which was applied to ultraHD PET was 6.08, 6.08, 4.45, 4.58%. It which was not applied to SharpIR was 5.93, 4.82, 4.45, 5.09%, it which was applied to SharpIR was 4.80, 3.92, 3.63, 4.50%. The resolution of phantom line of which location was upper, center, right, which was not applied to Astonish TF was 10.77, 11.54, 9.34 mm it which was applied to Astonish TF was 9.54, 8.90, 8.88 mm. It which was not applied to ultraHD PET was 7.84, 6.95, 8.32 mm, it which was applied to ultraHD PET was 7.51, 6.66, 8.27 mm. It which was not applied to SharpIR was 9.35, 8.69, 8.99, it which was applied to SharpIR was 9.88, 9.18, 9.00 mm. Conclusion : Image quality was advanced generally while reconstruction program which is based on time of flight was used. Futhermore difference of result compared each manufacture reconstruction program showed up, however this is caused by specification of instrument of each manufacture and difference of reconstruction algorithm. Therefore we need further examination to find out appropriate reconstruction condition while using reconstruction program used for advance of image quality.

• Progress in Medical Physics
• /
• v.22 no.3
• /
• pp.140-147
• /
• 2011

The purpose of this study was to estimate internal motion using molecular sieve for quantitative improvement of lung tumor and to localize lung tumor in the small animal PET image by evaluated data. Internal motion has been demonstrated in small animal lung region by molecular sieve contained radioactive substance. Molecular sieve for internal lung motion target was contained approximately 37 kBq Cu-64. The small animal PET images were obtained from Siemens Inveon scanner using external trigger system (BioVet). SD-Rat PET images were obtained at 60 min post injection of FDG 37 MBq/0.2 mL via tail vein for 20 min. Each line of response in the list-mode data was converted to sinogram gated frames (2~16 bin) by trigger signal obtained from BioVet. The sinogram data was reconstructed using OSEM 2D with 4 iterations. PET images were evaluated with count, SNR, FWHM from ROI drawn in the target region for quantitative tumor analysis. The size of molecular sieve motion target was $1.59{\times}2.50mm$. The reference motion target FWHM of vertical and horizontal was 2.91 mm and 1.43 mm, respectively. The vertical FWHM of static, 4 bin and 8 bin was 3.90 mm, 3.74 mm, and 3.16 mm, respectively. The horizontal FWHM of static, 4 bin and 8 bin was 2.21 mm, 2.06 mm, and 1.60 mm, respectively. Count of static, 4 bin, 8 bin, 12 bin and 16 bin was 4.10, 4.83, 5.59, 5.38, and 5.31, respectively. The SNR of static, 4 bin, 8 bin, 12 bin and 16 bin was 4.18, 4.05, 4.22, 3.89, and 3.58, respectively. The FWHM were improved in accordance with gate number increase. The count and SNR were not proportionately improve with gate number, but shown the highest value in specific bin number. We measured the optimal gate number what minimize the SNR loss and gain improved count when imaging lung tumor in small animal. The internal motion estimation provide localized tumor image and will be a useful method for organ motion prediction modeling without external motion monitoring system.

• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.161-177
• /
• 2019

In this paper, we study the performance improvement of the answer extraction in Question-Answering system by using sentence dependency parsing result. The Question-Answering (QA) system consists of query analysis, which is a method of analyzing the user's query, and answer extraction, which is a method to extract appropriate answers in the document. And various studies have been conducted on two methods. In order to improve the performance of answer extraction, it is necessary to accurately reflect the grammatical information of sentences. In Korean, because word order structure is free and omission of sentence components is frequent, dependency parsing is a good way to analyze Korean syntax. Therefore, in this study, we improved the performance of the answer extraction by adding the features generated by dependency parsing analysis to the inputs of the answer extraction model (Bidirectional LSTM-CRF). The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. In this study, we compared the performance of the answer extraction model when inputting basic word features generated without the dependency parsing and the performance of the model when inputting the addition of the Eojeol tag feature and dependency graph embedding feature. Since dependency parsing is performed on a basic unit of an Eojeol, which is a component of sentences separated by a space, the tag information of the Eojeol can be obtained as a result of the dependency parsing. The Eojeol tag feature means the tag information of the Eojeol. The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. From the dependency parsing result, a graph is generated from the Eojeol to the node, the dependency between the Eojeol to the edge, and the Eojeol tag to the node label. In this process, an undirected graph is generated or a directed graph is generated according to whether or not the dependency relation direction is considered. To obtain the embedding of the graph, we used Graph2Vec, which is a method of finding the embedding of the graph by the subgraphs constituting a graph. We can specify the maximum path length between nodes in the process of finding subgraphs of a graph. If the maximum path length between nodes is 1, graph embedding is generated only by direct dependency between Eojeol, and graph embedding is generated including indirect dependencies as the maximum path length between nodes becomes larger. In the experiment, the maximum path length between nodes is adjusted differently from 1 to 3 depending on whether direction of dependency is considered or not, and the performance of answer extraction is measured. Experimental results show that both Eojeol tag feature and dependency graph embedding feature improve the performance of answer extraction. In particular, considering the direction of the dependency relation and extracting the dependency graph generated with the maximum path length of 1 in the subgraph extraction process in Graph2Vec as the input of the model, the highest answer extraction performance was shown. As a result of these experiments, we concluded that it is better to take into account the direction of dependence and to consider only the direct connection rather than the indirect dependence between the words. The significance of this study is as follows. First, we improved the performance of answer extraction by adding features using dependency parsing results, taking into account the characteristics of Korean, which is free of word order structure and omission of sentence components. Second, we generated feature of dependency parsing result by learning - based graph embedding method without defining the pattern of dependency between Eojeol. Future research directions are as follows. In this study, the features generated as a result of the dependency parsing are applied only to the answer extraction model in order to grasp the meaning. However, in the future, if the performance is confirmed by applying the features to various natural language processing models such as sentiment analysis or name entity recognition, the validity of the features can be verified more accurately.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.1-21
• /
• 2020

With the rapid development of artificial intelligence technology, various techniques have been developed to extract meaningful information from unstructured text data which constitutes a large portion of big data. Over the past decades, text mining technologies have been utilized in various industries for practical applications. In the field of business intelligence, it has been employed to discover new market and/or technology opportunities and support rational decision making of business participants. The market information such as market size, market growth rate, and market share is essential for setting companies' business strategies. There has been a continuous demand in various fields for specific product level-market information. However, the information has been generally provided at industry level or broad categories based on classification standards, making it difficult to obtain specific and proper information. In this regard, we propose a new methodology that can estimate the market sizes of product groups at more detailed levels than that of previously offered. We applied Word2Vec algorithm, a neural network based semantic word embedding model, to enable automatic market size estimation from individual companies' product information in a bottom-up manner. The overall process is as follows: First, the data related to product information is collected, refined, and restructured into suitable form for applying Word2Vec model. Next, the preprocessed data is embedded into vector space by Word2Vec and then the product groups are derived by extracting similar products names based on cosine similarity calculation. Finally, the sales data on the extracted products is summated to estimate the market size of the product groups. As an experimental data, text data of product names from Statistics Korea's microdata (345,103 cases) were mapped in multidimensional vector space by Word2Vec training. We performed parameters optimization for training and then applied vector dimension of 300 and window size of 15 as optimized parameters for further experiments. We employed index words of Korean Standard Industry Classification (KSIC) as a product name dataset to more efficiently cluster product groups. The product names which are similar to KSIC indexes were extracted based on cosine similarity. The market size of extracted products as one product category was calculated from individual companies' sales data. The market sizes of 11,654 specific product lines were automatically estimated by the proposed model. For the performance verification, the results were compared with actual market size of some items. The Pearson's correlation coefficient was 0.513. Our approach has several advantages differing from the previous studies. First, text mining and machine learning techniques were applied for the first time on market size estimation, overcoming the limitations of traditional sampling based- or multiple assumption required-methods. In addition, the level of market category can be easily and efficiently adjusted according to the purpose of information use by changing cosine similarity threshold. Furthermore, it has a high potential of practical applications since it can resolve unmet needs for detailed market size information in public and private sectors. Specifically, it can be utilized in technology evaluation and technology commercialization support program conducted by governmental institutions, as well as business strategies consulting and market analysis report publishing by private firms. The limitation of our study is that the presented model needs to be improved in terms of accuracy and reliability. The semantic-based word embedding module can be advanced by giving a proper order in the preprocessed dataset or by combining another algorithm such as Jaccard similarity with Word2Vec. Also, the methods of product group clustering can be changed to other types of unsupervised machine learning algorithm. Our group is currently working on subsequent studies and we expect that it can further improve the performance of the conceptually proposed basic model in this study.

• Journal of Internet Computing and Services
• /
• v.14 no.5
• /
• pp.1-10
• /
• 2013

RFID(Radio Frequency Identification) system is non-contact identification technology. A basic RFID system consists of a reader, and a set of tags. RFID tags can be divided into active and passive tags. Active tags with power source allows their own operation execution and passive tags are small and low-cost. So passive tags are more suitable for distribution industry than active tags. A reader processes the information receiving from tags. RFID system achieves a fast identification of multiple tags using radio frequency. RFID systems has been applied into a variety of fields such as distribution, logistics, transportation, inventory management, access control, finance and etc. To encourage the introduction of RFID systems, several problems (price, size, power consumption, security) should be resolved. In this paper, we proposed an algorithm to significantly alleviate the collision problem caused by simultaneous responses of multiple tags. In the RFID systems, in anti-collision schemes, there are three methods: probabilistic, deterministic, and hybrid. In this paper, we introduce ALOHA-based protocol as a probabilistic method, and Tree-based protocol as a deterministic one. In Aloha-based protocols, time is divided into multiple slots. Tags randomly select their own IDs and transmit it. But Aloha-based protocol cannot guarantee that all tags are identified because they are probabilistic methods. In contrast, Tree-based protocols guarantee that a reader identifies all tags within the transmission range of the reader. In Tree-based protocols, a reader sends a query, and tags respond it with their own IDs. When a reader sends a query and two or more tags respond, a collision occurs. Then the reader makes and sends a new query. Frequent collisions make the identification performance degrade. Therefore, to identify tags quickly, it is necessary to reduce collisions efficiently. Each RFID tag has an ID of 96bit EPC(Electronic Product Code). The tags in a company or manufacturer have similar tag IDs with the same prefix. Unnecessary collisions occur while identifying multiple tags using Query Tree protocol. It results in growth of query-responses and idle time, which the identification time significantly increases. To solve this problem, Collision Tree protocol and M-ary Query Tree protocol have been proposed. However, in Collision Tree protocol and Query Tree protocol, only one bit is identified during one query-response. And, when similar tag IDs exist, M-ary Query Tree Protocol generates unnecessary query-responses. In this paper, we propose Adaptive M-ary Query Tree protocol that improves the identification performance using m-bit recognition, collision information of tag IDs, and prediction technique. We compare our proposed scheme with other Tree-based protocols under the same conditions. We show that our proposed scheme outperforms others in terms of identification time and identification efficiency.

• Journal of Intelligence and Information Systems
• /
• v.19 no.4
• /
• pp.147-157
• /
• 2013

With the advent of the digital environment that can be accessed anytime, anywhere with the introduction of high-speed network, the free distribution and use of digital content were made possible. Ironically this environment is raising a variety of copyright infringement, and product images used in the online shopping mall are pirated frequently. There are many controversial issues whether shopping mall images are creative works or not. According to Supreme Court's decision in 2001, to ad pictures taken with ham products is simply a clone of the appearance of objects to deliver nothing but the decision was not only creative expression. But for the photographer's losses recognized in the advertising photo shoot takes the typical cost was estimated damages. According to Seoul District Court precedents in 2003, if there are the photographer's personality and creativity in the selection of the subject, the composition of the set, the direction and amount of light control, set the angle of the camera, shutter speed, shutter chance, other shooting methods for capturing, developing and printing process, the works should be protected by copyright law by the Court's sentence. In order to receive copyright protection of the shopping mall images by the law, it is simply not to convey the status of the product, the photographer's personality and creativity can be recognized that it requires effort. Accordingly, the cost of making the mall image increases, and the necessity for copyright protection becomes higher. The product images of the online shopping mall have a very unique configuration unlike the general pictures such as portraits and landscape photos and, therefore, the general image watermarking technique can not satisfy the requirements of the image watermarking. Because background of product images commonly used in shopping malls is white or black, or gray scale (gradient) color, it is difficult to utilize the space to embed a watermark and the area is very sensitive even a slight change. In this paper, the characteristics of images used in shopping malls are analyzed and a watermarking technology which is suitable to the shopping mall images is proposed. The proposed image watermarking technology divide a product image into smaller blocks, and the corresponding blocks are transformed by DCT (Discrete Cosine Transform), and then the watermark information was inserted into images using quantization of DCT coefficients. Because uniform treatment of the DCT coefficients for quantization cause visual blocking artifacts, the proposed algorithm used weighted mask which quantizes finely the coefficients located block boundaries and coarsely the coefficients located center area of the block. This mask improves subjective visual quality as well as the objective quality of the images. In addition, in order to improve the safety of the algorithm, the blocks which is embedded the watermark are randomly selected and the turbo code is used to reduce the BER when extracting the watermark. The PSNR(Peak Signal to Noise Ratio) of the shopping mall image watermarked by the proposed algorithm is 40.7~48.5[dB] and BER(Bit Error Rate) after JPEG with QF = 70 is 0. This means the watermarked image is high quality and the algorithm is robust to JPEG compression that is used generally at the online shopping malls. Also, for 40% change in size and 40 degrees of rotation, the BER is 0. In general, the shopping malls are used compressed images with QF which is higher than 90. Because the pirated image is used to replicate from original image, the proposed algorithm can identify the copyright infringement in the most cases. As shown the experimental results, the proposed algorithm is suitable to the shopping mall images with simple background. However, the future study should be carried out to enhance the robustness of the proposed algorithm because the robustness loss is occurred after mask process.

• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.