• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.521-528
• /
• 2021

Existing image analysis systems in use in the military field are carried out by readers analyzing and identifying images themselves, writing and disseminating related content, and in this process, repetitive tasks are frequent, resulting in workload. In this paper, to solve the previous problem, we proposed an algorithm that can operate the Seq2Seq model on a word basis, which operates on a sentence basis, and applied the Attention technique to improve accuracy. In addition, by applying the Beam Search technique, we would like to recommend various current identification sentences based on the past identification contents of a specific area. It was confirmed through experiments that the Beam Search technique recommends sentences more effectively than the existing greedy Search technique, and confirmed that the accuracy of recommendation increases when the size of Beam is large.

• Journal of Digital Convergence
• /
• v.19 no.11
• /
• pp.463-470
• /
• 2021

Due to natural and human factors, dunhuang frescoes in China have gradually degenerated. China has conducted digital transformation of frescoes since 1990. However, it requires expensive research costs. Most of the visitors passively accept transformation contents and lack subjective participation. The paper focuses on produces a prototype of Dunhuang frescoes on the CoSpaces EDU. It was implemented as a metaverse through procedures such as transforming the cave into 3D, mapping images to the cave model and developing CoBlocks. The research puts forward a more specific methodology without expensive costs of development. The paper makes it easier to realize the immersive and interactive virtual Dunhuang frescoes world, to improve the tourism contents and educational effect. This research carries on the statistics to the product result which develops according to the user experience of 100 different ages, has obtained the good feedback. The research results need to be improved is to strengthen the processing of graphic details. Therefore, Optimization and improvement of the result will be carried out in the future research.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.491-500
• /
• 2021

In this study, we propose a method for detecting fall behavior using MS Kinect v2 RGBD Camera-based Human-Skeleton Keypoints and a 2-Stacked Bi-LSTM model. In previous studies, skeletal information was extracted from RGB images using a deep learning model such as OpenPose, and then recognition was performed using a recurrent neural network model such as LSTM and GRU. The proposed method receives skeletal information directly from the camera, extracts 2 time-series features of acceleration and distance, and then recognizes the fall behavior using the 2-Stacked Bi-LSTM model. The central joint was obtained for the major skeletons such as the shoulder, spine, and pelvis, and the movement acceleration and distance from the floor were proposed as features of the central joint. The extracted features were compared with models such as Stacked LSTM and Bi-LSTM, and improved detection performance compared to existing studies such as GRU and LSTM was demonstrated through experiments.

• Journal of Platform Technology
• /
• v.9 no.4
• /
• pp.60-70
• /
• 2021

This paper proposes a method capable of stably tracking the coordinates of a patterned laser image in a curved screen shooting system using two or more channels of multiple cameras. This method can track and acquire target points very effectively when applied to a multi-screen shooting method that can replace the HMD shooting method. Images of curved screens with severe deformation obtained from individual cameras are corrected through image normalization, image binarization, and noise removal. This corrected image is created and applied as an Euclidean space map that is easy to track the firing point based on the matching point. As a result of the experiment, the image coordinates of the pattern laser were stably extracted in the curved screen shooting system, and the error of the target point position of the real-world coordinate position and the broadband Euclidean map was minimized. The reliability of the proposed method was confirmed through the experiment.

• Journal of Korean Neurosurgical Society
• /
• v.65 no.4
• /
• pp.572-581
• /
• 2022

Objective : Compared to microscopes, exoscopes have advantages in field-depth, ergonomics, and educational value. Exoscopes are especially well-poised for adaptation into fluorescence-guided surgery (FGS) due to their excitation source, light path, and image processing capabilities. We evaluated the feasibility of near-infrared FGS using a 3-dimensional (3D), 4 K exoscope with near-infrared fluorescence imaging capability. We then compared it to the most sensitive, commercially-available near-infrared exoscope system (3D and 960 p). In-vitro and intraoperative comparisons were performed. Methods : Serial dilutions of indocyanine-green (1-2000 ㎍/mL) were imaged with the 3D, 4 K Olympus Orbeye (system 1) and the 3D, 960 p VisionSense Iridium (system 2). Near-infrared sensitivity was calculated using signal-to-background ratios (SBRs). In addition, three patients with brain tumors were administered indocyanine-green and imaged with system 1, with two also imaged with system 2 for comparison. Results : Systems 1 and 2 detected near-infrared fluorescence from indocyanine green concentrations of >250 ㎍/L and >31.3 ㎍/L, respectively. Intraoperatively, system 1 visualized strong near-infrared fluorescence from two, strongly gadolinium-enhancing meningiomas (SBR=2.4, 1.7). The high-resolution, bright images were sufficient for the surgeon to appreciate the underlying anatomy in the near-infrared mode. However, system 1 was not able to visualize fluorescence from a weakly-enhancing intraparenchymal metastasis. In contrast, system 2 successfully visualized both the meningioma and the metastasis but lacked high resolution stereopsis. Conclusion : Three-dimensional exoscope systems provide an alternative visualization platform for both standard microsurgery and near-infrared fluorescent guided surgery. However, when tumor fluorescence is weak (i.e., low fluorophore uptake, deep tumors), highly sensitive near-infrared visualization systems may be required.

• Korean Journal of Remote Sensing
• /
• v.38 no.5_1
• /
• pp.559-570
• /
• 2022

When a satellite image has low spatial resolution, it is difficult to detect small objects. In this research, we aim to check the effect of super resolution on object detection. Super resolution is a software method that increases the resolution of an image. Unpaired super resolution network is used to improve Sentinel-2's spatial resolution from 10 m to 3.2 m. Faster-RCNN, RetinaNet, FCOS, and S²ANet were used to detect vessels in the Sentinel-2 images. We experimented the change in vessel detection performance when super resolution is applied. As a result, the Average Precision (AP) improved by at least 12.3% and up to 33.3% in the ship detection models trained with the super-resolution image. False positive and false negative cases also decreased. This implies that super resolution can be an important pre-processing step in object detection, and it is expected to greatly contribute to improving the accuracy of other image-based deep learning technologies along with object detection.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.35 no.5
• /
• pp.259-266
• /
• 2022

In this study, we developed a deep learning and vision sensor-based assembly performance evaluation method isfor prefabricated steel structures. The assembly parts were segmented using a modified version of the receptive field block convolution module inspired by the eccentric function of the human visual system. The quality of the assembly was evaluated by detecting the bolt holes in the segmented assembly part and calculating the bolt hole positions. To validate the performance of the evaluation, models of standard and defective assembly parts were produced using a 3D printer. The assembly part segmentation network was trained based on the 3D model images captured from a vision sensor. The sbolt hole positions in the segmented assembly image were calculated using image processing techniques, and the assembly performance evaluation using the k-nearest neighbor algorithm was verified. The experimental results show that the assembly parts were segmented with high precision, and the assembly performance based on the positions of the bolt holes in the detected assembly part was evaluated with a classification error of less than 5%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.7
• /
• pp.910-916
• /
• 2021

In modern society, various digital communication equipment is being used due to the influence of the 4th industrial revolution. Accordingly, interest in removing noise generated in a data transmission process is increasing, and research is being conducted to efficiently reconstruct an image. In this paper, we propose a filtering algorithm to remove the AWGN generated in the digital image transmission process. The proposed algorithm classifies pixels with high similarity by selecting regions with similar patterns around the input pixels according to block matching to remove the AWGN that appears strongly in the image. The selected pixel determines the estimated value by applying the weight obtained by the local steering kernel, and obtains the final output by adding or subtracting the input pixel value according to the standard deviation of the center mask. In order to evaluate the proposed algorithm, it was simulated with existing AWGN removal algorithms, and comparative analysis was performed using enlarged images and PSNR.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.4
• /
• pp.17-22
• /
• 2022

With the spread of web services, data types are becoming more diversified. In addition to the form of storing data such as images, videos, and texts, the number and form of properties and metadata expressing the data are different for each data. In order to efficiently process such unstructured data, a key-value store is widely used for state-of-the-art applications. LSM-Tree (Log Structured Merge Tree) is the core data structure of various commercial key-value stores. LSM-Tree is optimized to provide high performance for small writes by recording all write and delete operations in a log manner. However, there is a problem in that the delay time and processing speed of user requests are lowered as batches of deletion operations for expired data are inserted into the LSM-Tree as special key-value data. This paper presents a Filtered LSM-Tree (FLSM-Tree) that solves the above problem by separating the deleted key from the main tree structure while maintaining all the advantages of the existing LSM-Tree. The proposed method is implemented in LevelDB, a commercial key-value store and it shows that the read performance is improved by up to 47% in performance evaluation.

• Journal of Korea Multimedia Society
• /
• v.25 no.5
• /
• pp.748-756
• /
• 2022

Thanks to the competition of AlphaGo and Sedol Lee, machine learning has received world-wide attention and huge investments. The performance improvement of computing devices greatly contributed to big data processing and the development of neural networks. Artificial intelligence not only imitates human beings in many fields, but also seems to be better than human capabilities. Although humans' creation is still considered to be better and higher, several artificial intelligences continue to challenge human creativity. The quality of some creative outcomes by AI is as good as the real ones produced by human beings. Sometimes they are not distinguishable, because the neural network has the competence to learn the common features contained in big data and copy them. In order to confirm whether artificial intelligence can express the inherent characteristics of different arts, this paper proposes a new neural network model called Humming. It is an experimental model that combines vgg16, which extracts image features, and DeepJ's architecture, which excels in creating various genres of music. A dataset produced by our experiment shows meaningful and valid results. Different results, however, are produced when the amount of data is increased. The neural network produced a similar pattern of music even though it was a different classification of images, which was not what we were aiming for. However, these new attempts may have explicit significance as a starting point for feature transfer that will be further studied.