• Smart Media Journal
• /
• v.9 no.4
• /
• pp.17-25
• /
• 2020

In the image object classification problem, low-resolution images may have a negative impact on the classification result, especially when the classification method, such as a convolutional neural network (CNN) model, is trained on a high-resolution (HR) image dataset. In this paper, we analyze the behavior of applying a classical super-resolution (SR) method such as bicubic interpolation, and a deep CNN model such as SRCNN to enhance low-resolution (LR) weeds images used for classification. Using an HR dataset, we first train a CNN model for weeds image classification with a default input size of 128 × 128. Then, given an LR weeds image, we rescale to default input size by applying the bicubic interpolation or the SRCNN model. We analyze these two approaches on the Chonnam National University (CNU) weeds dataset and find that SRCNN is suitable for the image size is smaller than 80 × 80, while bicubic interpolation is convenient for a larger image.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.98-101
• /
• 2019

At present, deep convolutional neural networks have made a very important contribution in single-image super-resolution. Through the learning of the neural networks, the features of input images are transformed and combined to establish a nonlinear mapping of low-resolution images to high-resolution images. Some previous methods are difficult to train and take up a lot of memory. In this paper, we proposed a simple and compact deep recursive residual network learning the features for single image super resolution. Global residual learning and local residual learning are used to reduce the problems of training deep neural networks. And the recursive structure controls the number of parameters to save memory. Experimental results show that the proposed method improved image qualities that occur in previous methods.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.63-66
• /
• 2018

Recently, a convolutional neural network (CNN) models at single image super-resolution have been very successful. Residual learning improves training stability and network performance in CNN. In this paper, we compare four convolutional neural network models for super-resolution (SR) to learn nonlinear mapping from low-resolution (LR) input image to high-resolution (HR) target image. Four models include general CNN model, global residual learning CNN model, local residual learning CNN model, and the CNN model with global and local residual learning. Experiment results show that the results are greatly affected by how skip connections are connected at the basic CNN network, and network trained with only global residual learning generates highest performance among four models at objective and subjective evaluations.

• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.104-123
• /
• 2022

Style transfer based on neural network provides very high quality results by reflecting the high level structural characteristics of images, and thereby has recently attracted great attention. This paper deals with the problem of resolution limitation due to GPU memory in performing such neural style transfer. We can expect that the gradient operation for style transfer based on partial image, with the aid of the fixed size of receptive field, can produce the same result as the gradient operation using the entire image. Based on this idea, each component of the style transfer loss function is analyzed in this paper to obtain the necessary conditions for partitioning and padding, and to identify, among the information required for gradient calculation, the one that depends on the entire input. By structuring such information for using it as auxiliary constant input for partition-based gradient calculation, this paper develops a recursive algorithm for super high-resolution image style transfer. Since the proposed method performs style transfer by partitioning input image into the size that a GPU can handle, it can perform style transfer without the limit of the input image resolution accompanied by the GPU memory size. With the aid of such super high-resolution support, the proposed method can provide a unique style characteristics of detailed area which can only be appreciated in super high-resolution style transfer.

• Korean Journal of Remote Sensing
• /
• v.17 no.1
• /
• pp.45-56
• /
• 2001

To acquire seabed information, the mosaic images of the seabed were generated using Side Scan Sonar. Short time energy function which is needed for slant range correction is proposed to get the height of Tow-Fish to the reflected acoustic amplitudes of each ping, and that leads to a mosaic image without water column. While generating mosaic image, maximum value, last value and average value are used for the measure of a pixel in the mosaic image and 3-D information was kept by using acoustic amplitudes which were heading for specific direction. As a generating method of mosaic image, low resolution mosaic image which is over 1m/pixel resolution was generated for whole survey area first, and then high resolution mosaic image which is generated under 0.1m/pixel resolution was generated for the selected area. Rocks, ripple mark, sand wave, tidal flat and artificial fish reef are found in the mosaic image.

• Korean Journal of Remote Sensing
• /
• v.32 no.6
• /
• pp.551-566
• /
• 2016

In the satellite remote sensing, the operational environment of the satellite sensor causes image degradation during the image acquisition. The degradation results in noise and blurring which badly affect identification and extraction of useful information in image data. Especially, the degradation gives bad influence in the analysis of images collected over the scene with complicate surface structure such as urban area. This study proposes a multi-stage image restoration to improve the accuracy of detailed analysis for the images collected over the complicate scene. The proposed method assumes a Gaussian additive noise, Markov random field of spatial continuity, and blurring proportional to the distance between the pixels. Point-Jacobian Iteration Maximum A Posteriori (PJI-MAP) estimation is employed to restore a degraded image. The multi-stage process includes the image segmentation performing region merging after pixel-linking. A dissimilarity coefficient combining homogeneity and contrast is proposed for image segmentation. In this study, the proposed method was quantitatively evaluated using simulation data and was also applied to the two panchromatic images of super-high resolution: Dubaisat-2 data of 1m resolution from LA, USA and KOMPSAT3 data of 0.7 m resolution from Daejeon in the Korean peninsula. The experimental results imply that it can improve analytical accuracy in the application of remote sensing high resolution panchromatic imagery.

• Proceedings of the KIEE Conference
• /
• 1987.07b
• /
• pp.1402-1404
• /
• 1987

Photoconductive target was fabricated to make vidicon available. In order for a vidicon to operate well, it is essential that the target have high photosensitivity, low image lag, and high resolution. In the vidicon mode analysis, photosensitivity of 0.8, image lag of 30%, resolution of 300 TV lines, and the S/N ratio of 30 dB at 10 lux illumination were measured.

• Journal of Broadcast Engineering
• /
• v.25 no.2
• /
• pp.242-251
• /
• 2020

Recently, many deep convolutional neural networks for image super-resolution have been studied. Existing deep learning-based super-resolution algorithms are architecture that up-samples the resolution at the end of the network. The post-upsampling architecture has an inefficient structure at large scaling factor result of predicting a lot of information for mapping from low-resolution to high-resolution at once. In this paper, we propose a single image super-resolution using Channel Attention Residual Dense Block based on an iterative up-down sampling architecture. The proposed algorithm efficiently predicts the mapping relationship between low-resolution and high-resolution, and shows up to 0.14dB performance improvement and enhanced subjective image quality compared to the existing algorithm at large scaling factor result.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.35 no.1
• /
• pp.1-10
• /
• 2017

An UAV (Unmanned Aerial Vehicle) is a flight system that is designed to conduct missions without a pilot. Compared to traditional airborne-based photogrammetry, UAV-based photogrammetry is inexpensive and can obtain high-spatial resolution data quickly. In this study, we aimed to classify the land cover using high-spatial resolution images obtained using a UAV. An RGB camera was used to obtain high-spatial resolution orthoimage. For accurate classification, multispectral image about same areas were obtained using a multispectral sensor. A DSM (Digital Surface Model) and a modified NDVI (Normalized Difference Vegetation Index) were generated using images obtained using the RGB camera and multispectral sensor. Pixel-based classification was performed for twelve classes by using the RF (Random Forest) method. The classification accuracy was evaluated based on the error matrix, and it was confirmed that the proposed method effectively classified the area compared to supervised classification using only the RGB image.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.502-504
• /
• 2003

The main objective of this study is to examine the effectiveness of newly available high spatial resolution satellite images, in evaluating vegetation environment of the urban areas. In doing so, we have used satellite images from QuickBird and selected some areas of Fukuoka City, Kyushu Japan, as study area. The results of the study revealed that, high resolution images are more effective in close monitoring of the vegetation status and green plants should be planted in open spaces and roofs of urban areas to increase vegetation, which will in turn act as a remedy to reduce heat island phenomenon.