• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.8
• /
• pp.81-88
• /
• 2017

This paper proposes a method of detecting rain regions from a single image. More specifically, a way of training the deep convolutional neural network based on the collected rain and non-rain patches is presented in a supervised manner. It is also shown that the proposed rain detection method based on deep convolutional neural network can provide better performance than the conventional rain detection method based on dictionary learning. Moreover, it is confirmed that the application of the proposed rain detection for rain removal can lead to some improvement in detail representation on the low-frequency regions of the rain-removed images. Additionally, this paper introduces the rain transfer method that inserts rain patterns into original images, thereby producing rain effects on the resulting images. The proposed rain transfer method could be used to augment rain patterns while constructing rain database.

• Journal of Broadcast Engineering
• /
• v.20 no.1
• /
• pp.153-163
• /
• 2015

This paper proposes a rain detection and removal algorithm that is robust against camera motion in video sequences. In detection part, the proposed algorithm initially detects possible rain streaks by using intensity properties and spatial properties. Then, the rain streak candidates are selected based on Gaussian distribution model. In removal part, a non-rain block matching algorithm is performed between adjacent frames to find similar blocks to the block that has rain pixels. If the similar blocks to the block are obtained, the rain region of the block is reconstructed by non-local means (NLM) filter using the similar neighbors. Experimental results show that the proposed algorithm outperforms the previous works in terms of subjective visual quality of de-rained video sequences.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.9
• /
• pp.96-105
• /
• 2015

This paper proposes a DSP optimization solution of rain detection and removal algorithm. We propose rain detection and removal algorithms considering camera motion, and also presents optimization results in algorithm level and DSP level. At algorithm level, this paper utilizes a block level binary pattern analysis, and reduces the operation time by using the fast motion estimation algorithm. Also, the algorithm is optimized at DSP level through inter memory optimization, EDMA, and software pipelining for real-time operation. Experiment results show that the proposed algorithm is superior to the other algorithms in terms of visual quality as well as processing speed.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.9 no.6
• /
• pp.345-351
• /
• 2014

This paper presents pedestrian recognition to improve performance for vehicle safety system or surveillance system. Pedestrian detection method using HOG (Histograms of Oriented Gradients) has showed 90% recognition rate. But if someone takes a picture in the rain, the image may be distorted by rain streaks and recognition rate goes down by 62%. To solve this problem, we applied image decomposition method using MCA (Morphological Component Analysis). In this case, rain removal method improves recognition rate from 62% to 70%. However, it is difficult to apply conventional image decomposition method using MCA on vehicle safety system or surveillance system as conventional method is too slow for real-time system. To alleviate this issue, we propose a rain removal method by using low-pass filter and DCT (Discrete Cosine Transform). The DCT helps separate the image into rain components. The image is removed rain components by Butterworth filtering. Experimental results show that our method achieved 90% of recognition rate. In addition, the proposed method had accelerated processing time to 17.8ms which is acceptable for real-time system.

• Journal of electromagnetic engineering and science
• /
• v.13 no.3
• /
• pp.151-157
• /
• 2013

Marine radars are affected by sea and rain clutters, which can make target discrimination difficult. The clutter standard deviation and improvement factor are applied using multiple parameters-moving speed of radar, antenna speed, angle, etc. When a radar signal is processed, a Data Matrix Bank (DMB) filter can be applied to remove sea clutters. This filter allows detection of a target, and since it is not affected by changes in adjacent clutters resulting from a multi- target signal, sea state clutters can be removed. In this paper, we study the level for clutter removal and the method for target detection. In addition, we design a signal processing algorithm for marine radars, analyze the performance of the DMB filter algorithm, and provide a DMB filter algorithm design. We also perform a DMB filter algorithm analysis and simulation, and then apply this to the DMB filter and cell-average constant false alarm rate design to show comparative results.

• Journal of Sensor Science and Technology
• /
• v.26 no.3
• /
• pp.149-154
• /
• 2017

Detection of rain is one of the essential weather factors that are monitored by automatic weather stations in Korea. In this work, we studied the operation standards required for impedance-based rain detectors in terms of surface temperature and sensitivity, in an effort to establish a qualification procedure for rain detectors. The surface temperature of rain detectors was measured at varying air temperatures from $-30^{\circ}C$ to $20^{\circ}C$, considering the hypothetical presence and absence of rain/snow. In addition, the sensitivity of rain detectors was studied generating artificial raindrops of regular size. The sensitivity was evaluated in terms of the critical number of droplets that triggers the activation of the rain detector. We found that the sensitivity is affected by stationary, horizontal, and vertical droplet deposition methods. The critical number of droplets for the stationary deposition is higher than that for both horizontal and vertical depositions, which provides the maximum limit of droplets required to activate the detector. Based on our experiments considering surface temperature measurements and sensitivity tests, we suggest a revised version of surface temperature and sensitivity requirements for the qualification of impedance-based rain detectors.

• Bulletin of the Korean Chemical Society
• /
• v.26 no.12
• /
• pp.1937-1940
• /
• 2005

A lab-made chemiluminescence system with air pump was developed for monitoring of some metal levels in rain. The air pump enabled injection of 17.7 $\mu$g samples into a glass cell filled with luminol-$H_2O_2$ reagent of typically 300 $\mu$L for chemiluminescence measurement. The monitored trend of total metal ions in the rain collected in our campus was compared with analytical results of each metal ion from GFAAS. The system was also demonstrated to determine $Cr^{6+}$ by reduction to $Cr^{3+}$ using $SnCl_2$. The limit of detection for $Cr^{6+}$ obtained by 4 measurements was 85.0 pg $mL^{-1}$ with a relative standard deviation of 3.4%. Although this system doesn’t have selectivity due to the characteristics of chemiluminescence, application of it to environmental monitoring as a sensor for some transition metal ions was demonstrated.

• Journal of Korea Water Resources Association
• /
• v.53 no.4
• /
• pp.313-322
• /
• 2020

There is always a risk of water disasters due to sudden storms in mountainous regions in Korea, which is more than 70% of the country's land. In this study, a radar-based risk prediction technique for sudden downpour is applied in the mountainous region and is evaluated for its applicability using Mt. Biseul rain radar. Eight local heavy rain events in mountain regions are selected and the information was calculated such as early detection of cumulonimbus convective cells, automatic detection of convective cells, and risk index of detected convective cells using the three-dimensional radar reflectivity, rainfall intensity, and doppler wind speed. As a result, it was possible to confirm the initial detection timing and location of convective cells that may develop as a localized heavy rain, and the magnitude and location of the risk determined according to whether or not vortices were generated. In particular, it was confirmed that the ground rain gauge network has limitations in detecting heavy rains that develop locally in a narrow area. Besides, it is possible to secure a time of at least 10 minutes to a maximum of 65 minutes until the maximum rainfall intensity occurs at the time of obtaining the risk information. Therefore, it would be useful as information to prevent flash flooding disaster and marooned accidents caused by heavy rain in the mountainous area using this technique.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.113-117
• /
• 2020

Images were taken under various weather such as rain, haze, snow often show low visibility, which can dramatically decrease accuracy of some tasks in computer vision: object detection, segmentation. Besides, previous work to enhance image usually downsample the image to receive consistency features but have not yet good upsample algorithm to recover original size. So, in this research, we jointly implement removal streak in heavy rain image and super resolution using a deep network. We put forth a 2-stage network: a multi-model network followed by a refinement network. The first stage using rain formula in the single image and two operation layers (addition, multiplication) removes rain streak and noise to get clean image in low resolution. The second stage uses refinement network to recover damaged background information as well as upsample, and receive high resolution image. Our method improves visual quality image, gains accuracy in human action recognition task in datasets. Extensive experiments show that our network outperforms the state of the art (SoTA) methods.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.12 no.4
• /
• pp.44-55
• /
• 2013

Intelligent transportation systems allow us to have valuable opportunities for collecting reliable wide-area coverage traffic and weather data. Significant efforts have been made in many countries to apply these data. This study identifies the critical points for classifying rain intensity by analyzing the relationship between rainfall and the amount of speed reduction. Then, traffic prediction performance by rain intensity level is evaluated using relative errors. The results show that critical points are 0.4mm/5min and 0.8mm/5min for classifying rain intensity (slight, moderate, and heavy rain). The best prediction performance is observable when previous five-block speed data is used as inputs under normal weather conditions. On the other hand, previous two or three-block speed data is used as inputs under rainy weather conditions. The outcomes of this study support the development of more reliable traffic information for providing advanced traffic information service.