• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.1
• /
• pp.19-26
• /
• 2022

This paper presents a pre-processing method to facilitate the container's BIC-code recognition. We propose a network that can find ROI(Region Of Interests) containing a BIC-code region and estimate a homography matrix for warping. Taking the structure of STN(Spatial Transformer Networks), the proposed network consists of next 3 steps, ROI detection, homography matrix estimation, and warping using the homography estimated in the previous step. It contributes to improving the accuracy of BIC-code recognition by estimating ROI and matrix using the proposed network and correcting perspective distortion of ROI using the estimated matrix. For performance evaluation, five evaluators evaluated the output image as a perfect score of 5 and received an average of 4.25 points, and when visually checked, 224 out of 312 photos are accurately and perfectly corrected, containing ROI.

• Journal of Cadastre & Land InformatiX
• /
• v.48 no.2
• /
• pp.171-183
• /
• 2018

Despite the rapid expansion of satellite images supply, the application of imagery is often restricted due to unautomated image processing. This paper presents the automated process for the selection of training areas which are essential to conducting supervised image classification. The training areas were selected based on the prior and cover information. After the selection, the training data were used to classify land cover in an urban area with the latest image and the classification accuracy was valuated. The automatic selection of training area was processed with following steps, 1) to redraw inner areas of prior land cover polygon with negative buffer (-15m) 2) to select the polygons with proper size of area ($2,000{\sim}200,000m^2$) 3) to calculate the mean and standard deviation of reflectance and NDVI of the polygons 4) to select the polygons having characteristic mean value of each land cover type with minimum standard deviation. The supervised image classification was conducted using the automatically selected training data with Sentinel-2 images in 2017. The accuracy of land cover classification was 86.9% ($\hat{K}=0.81$). The result shows that the process of automatic selection is effective in image processing and able to contribute to solving the bottleneck in the application of imagery.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.9
• /
• pp.2121-2128
• /
• 2015

In this paper presents a method of extracting reliable and regions of interest (ROI) in barcode for the purpose of factory automation. backgrounds are separated based on directional components and the characteristics of detected patterns. post-processing is performed on candidate images with analysis of problems caused by blur, rotation and areas of high similarity. In addition, the resizing factor is used to achieve faster calculations through image resizing. The input images contained multiple product or barcode for application to diverse automation environments; a high extraction success rate is accomplished despite the maximum shooting distance of 80 cm. Simulations involving images with various shooting distances gave an ROI detection rate of 100% and a post-processing success rate of 99.3%.

• Journal of The Korean Association For Science Education
• /
• v.27 no.9
• /
• pp.893-906
• /
• 2007

The purposes of this study are to understand pre-service science teachers' areas of practice concern and reflections on the science classes during student-teaching, and to grasp the factors affecting their concerns and reflections. Four pre-service science teachers participated in this study. Data were collected from four pre-service science teachers' practice journals, instructional materials, and semi-structured individual interviews after their student-teaching. The results are as follows: firstly, the pre-service teachers' concern is focused on the teaching environment and strategies, particularly classroom atmosphere and class management. On the other hand, they pay little attention to science content. Secondly, pre-service teachers' reflections are confined within the limited areas such as classroom management, the proper role as science teachers, or various teaching materials. The level of their reflections is low, mostly concentrated on 'routine' or 'technical' level. Higher levels of reflections, such as 'dialogic' or 'transformative' are not revealed at all. Thirdly, the mentor teachers have the biggest influence upon the concerns and reflections of pre-service teachers during student-teaching.

• Journal of Broadcast Engineering
• /
• v.22 no.6
• /
• pp.765-773
• /
• 2017

Ultra-high definition videos, panorama contents, and ultra-wide viewing videos have a huge spatial resolution. However, a whole region of images is not always interesting to viewers due to limitations of system resources and display devices. To allow one or more interesting tiles to be decoded freely without decoding other tiles, this paper introduces a tile-based independent coding method. The propose method restraints motion vectors to be within a tile region shifting motion search area and modifying an initial motion vector. Experiments results show that any desired tile is capable to be decoded independently of other tiles.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.5
• /
• pp.137-151
• /
• 2014

This paper proposes a method of detecting object of interest(OOI) in general natural images. OOI is subjectively estimated by human in images. The vision of human, in general, might focus on OOI. As the first step for automatic detection of OOI, candidate regions of OOI are detected by using a saliency map based on the human visual perception. A saliency map locates an approximate OOI, but there is a problem that they are not accurately segmented. In order to address this problem, in the second step, an exact object region is automatically detected by combining graph-based image segmentation and skeletonization. In this paper, we calculate the precision, recall and accuracy to compare the performance of the proposed method to existing methods. In experimental results, the proposed method has achieved better performance than existing methods by reducing the problems such as under detection and over detection.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.189-198
• /
• 2024

Lipreading is one of the important parts of speech recognition, and several studies have been conducted to improve the performance of lipreading in lipreading systems for speech recognition. Recent studies have used method to modify the model architecture of lipreading system to improve recognition performance. Unlike previous research that improve recognition performance by modifying model architecture, we aim to improve recognition performance without any change in model architecture. In order to improve the recognition performance without modifying the model architecture, we refer to the cues used in human lipreading and set other regions such as chin and cheeks as regions of interest along with the lip region, which is the existing region of interest of lipreading systems, and compare the recognition rate of each region of interest to propose the highest performing region of interest In addition, assuming that the difference in normalization results caused by the difference in interpolation method during the process of normalizing the size of the region of interest affects the recognition performance, we interpolate the same region of interest using nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation, and compare the recognition rate of each interpolation method to propose the best performing interpolation method. Each region of interest was detected by training an object detection neural network, and dynamic time warping templates were generated by normalizing each region of interest, extracting and combining features, and mapping the dimensionality reduction of the combined features into a low-dimensional space. The recognition rate was evaluated by comparing the distance between the generated dynamic time warping templates and the data mapped to the low-dimensional space. In the comparison of regions of interest, the result of the region of interest containing only the lip region showed an average recognition rate of 97.36%, which is 3.44% higher than the average recognition rate of 93.92% in the previous study, and in the comparison of interpolation methods, the bilinear interpolation method performed 97.36%, which is 14.65% higher than the nearest neighbor interpolation method and 5.55% higher than the bicubic interpolation method. The code used in this study can be found a https://github.com/haraisi2/Lipreading-Systems.

• Journal of Broadcast Engineering
• /
• v.14 no.3
• /
• pp.366-377
• /
• 2009

In this paper, we implement the real-time system which extracts 3-dimensional coordinates from right and left images captured by a stereo camera and provides users with reality through a virtual space operated by the 3-dimensional coordinates. In general, all pixels in correspondence region are compared for the disparity estimation. However, for a real time process, the central coordinates of the correspondence region are only used in the proposed algorithm. In the implemented system, 3D coordinates are obtained by using the depth information derived from the estimated disparity and we set user's hand as a region of interest(ROI). After user's hand is detected as the ROI, the system keeps tracking a hand's movement and generates a virtual space that is controled by the hand. Experimental results show that the implemented system could estimate the disparity in real -time and gave the mean-error less than 0.68cm within a range of distance, 1.5m. Also It had more than 90% accuracy in the hand recognition.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.6
• /
• pp.321-328
• /
• 2017

We have implemented a technique to correct the brightness, saturation, and contrast of an image according to the degree of light, and further compensate the backlight. Backlight compensation can be done automatically or manually. For manual backlight compensation, we have to select the region of interest (ROI). ROI can be selected by connecting the outline of the desired object. We make users select the region delicately with the new magnetic lasso tool. The previous lasso tool has a disadvantage that the start point and the end point must be connected. However, the proposed lasso tool has the advantage of selecting the region of interest without connecting the start point and the end point. We can automatically obtain various results of backlight compensation by adjusting the number of k-means clusters for texture extraction and the threshold value for binarization.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.7
• /
• pp.1628-1633
• /
• 2014

In this paper, we propose a news article recommendation system that reflects users' areas of interest and reading habits. Users can select interesting subject then our proposed system displays interesting articles above the other articles. Also the proposed system reflects users' dynamic interests using analyse of user's reading habits. The method of dynamic interest applies the different weight values from users simply clicking and reading entire articles. When users read articles from specific areas, the prosed system increases the weight of these specific areas using XML structure information. Hence users can read their articles of interest with ease.