• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.59-68
• /
• 2021

As artificial intelligence(AI) technologies, which have made rapid growth recently, began to be applied to the marine environment such as ships, there have been active researches on the application of CNN-based models specialized for digital videos. In E-Navigation service, which is combined with various technologies to detect floating objects of clash risk to reduce human errors and prevent fires inside ships, real-time processing is of huge importance. More functions added, however, mean a need for high-performance processes, which raises prices and poses a cost burden on shipowners. This study thus set out to propose a method capable of processing information at a high rate while maintaining the accuracy by applying Quantization techniques of a deep learning model. First, videos were pre-processed fit for the detection of floating matters in the sea to ensure the efficient transmission of video data to the deep learning entry. Secondly, the quantization technique, one of lightweight techniques for a deep learning model, was applied to reduce the usage rate of memory and increase the processing speed. Finally, the proposed deep learning model to which video pre-processing and quantization were applied was applied to various embedded boards to measure its accuracy and processing speed and test its performance. The proposed method was able to reduce the usage of memory capacity four times and improve the processing speed about four to five times while maintaining the old accuracy of recognition.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.10a
• /
• pp.224-226
• /
• 2016

With the technical development and rapid increase of private demand, the new market for unmanned vehicle combined with the characteristics of 'unmanned automation' and 'vehicle' is rapidly growing. Even though the pilot driving is currently allowed in some countries, there is no country that has institutionalized the formal driving of self-driving cars. In case of the existing vehicles, safety incidents are frequently happening due to the frequent malfunction of the rear sensor, blind spot of the rear camera, or drivers' carelessness. Once such minor flaws are complemented, the relevant regulations for the commercialization of self-driving car and small drone could be relieved. Contrary to the ultrasonic and laser sensors used for the existing vehicles, this paper aims to attempt the distance measurement by using the depth sensor. A depth camera calculates the distance data based on the TOF method calculating the time difference by lighting laser or infrared light onto an object or area and then receiving the beam coming back. As this camera can obtain the depth data in the pixel unit of CCD camera, it can be used for collecting depth data in real-time. This paper suggests to solve problems mentioned above by using depth data in real-time and also to design the obstacle avoidance system through distance measurement.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.1099-1110
• /
• 2023

The recognition system of autonomous driving and robot navigation performs vision work such as object recognition, tracking, and lane detection after multi-sensor fusion to improve performance. Currently, research on a deep learning model based on the fusion of a camera and a lidar sensor is being actively conducted. However, deep learning models are vulnerable to adversarial attacks through modulation of input data. Attacks on the existing multi-sensor-based autonomous driving recognition system are focused on inducing obstacle detection by lowering the confidence score of the object recognition model.However, there is a limitation that an attack is possible only in the target model. In the case of attacks on the sensor fusion stage, errors in vision work after fusion can be cascaded, and this risk needs to be considered. In addition, an attack on LIDAR's point cloud data, which is difficult to judge visually, makes it difficult to determine whether it is an attack. In this study, image scaling-based camera-lidar We propose an attack method that reduces the accuracy of LCCNet, a fusion model (camera-LiDAR calibration model). The proposed method is to perform a scaling attack on the point of the input lidar. As a result of conducting an attack performance experiment by size with a scaling algorithm, an average of more than 77% of fusion errors were caused.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.375-378
• /
• 2007

This paper describes indoor location estimation intelligent robot. It is loaded indoor location estimation function using RSSI based indoor location estimation system and wireless sensor networks. Spartan III(Xilinx, U.S.A.) is used as a main control device in the mobile robot and the current direction data is collected in the indoor location estimation system. The data is transferred to the wireless sensor network node attached to the mobile robot through Zigbee/IEEE 802.15.4, a wireless communication. After receiving it, with the data of magnetic compass the node is aware of and senses the direction the robot head for and the robot moves to its destination. Indoor location estimation intelligent robot is can be moved efficiently and actively without obstacle on flat ground to the appointment position by user.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.4
• /
• pp.675-683
• /
• 1992

Mapping and navigation system based on certainty grids for an autonomous mobile robt operating in unknown and unstructured environment is described. The system uses sonar range data to build a map of robot's surroundings. The range data from sonar sensor are integrated into a probability map that is composed of two dimensional grids which contain the probabilities of being occupied by the objects in the environment. A Bayesian model is used to estimate the uncertainty of the sensor information and to update the existing probability map with new range data. The resulting two dimensional map is used for path planning and navigation. In this paper, the Bayesian updating model which was successfully simulated in our earlier work is implemented on a mobile robot and is shown to be valid in the real world through experiment. This paper also proposes a technique for reducing for reducing specular reflection problem of sonar system which seriousely deteriorates the map quality, and a new path planning method based on weighted distance, which enables the robot to efficiently navigate in an unknown area.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.2
• /
• pp.313-320
• /
• 2018

The importance of sensors on a self-driving vehicle has rising since it act as eyes for the vehicle. Lidar sensors based on laser technology tend to yield better image quality with more laser channels, thus, it has higher detection accuracy for obstacles, pedistrians, terrain, and other vechicles. However, incorporating more laser channels results higher unit price more than ten times, and this is a major drawback for using high channel lidar sensors on a vehicle for actual consumer market. To come up with this drawback, we propose a method of integrating multiple low channel, low cost lidar sensors acting as one high channel sensor. The result uses four 16 channels lidar sensors with small form factor acting as one bulky 64 channels sensor, which in turn, improves vehicles cosmetic aspects and helps widespread of using the lidar technology for the market.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.332.1-332.1
• /
• 2014

정보통신 기술의 발전으로 지능형 자동차와 같은 미래형 고부가 가치 자동차 산업은 지속적인 성장이 기대된다. 그리고, 안전과 직결되는 차간 거리 계측은 다양한 종류의 센서에 의해 측정이 되고 있으며, 운전자 및 탑승자의 생명을 보호하고 있다. 다양한 차간 거리센서 중에서도 전방의 물체 인식 및 넓은 영역, 장거리에 대한 센싱은 레이저를 이용하여 구현할 수 있다. 본 논문은 자동차 뿐만 아니라, 지능형 자율 주행시 전방의 물체 인식 및 거리 계측 센서로 적용이 가능한 반도체 레이저의 설계 및 제작에 관해 소개한다. 반도체 레이저는 물질에 따라 각각 다른 파장대역의 광을 조사하고 있으며 이 레이저 빔은 물체에 맞고 부딪히면 반사되어 되돌아 오는 특성을 가지고 있다. 따라서, 펄스 구동에 의해 단위 펄스당 출사되는 레이저는 전방 물체에 부딪혀 되돌아 오는 시간을 구하게 되면 레이저 광원에서 물체까지의 거리를 구할 수가 있게 된다. 여기서 펄스 레이저의 출력은 물체 감지가 가능한 거리의 정보를 가지고 있으며, 펄스로 구동될 때 반복 주파수 및 펄스 폭은 각각 거리계측 시간과 분해능을 결정하는 주요 요소가 된다. 따라서, 장거리 물체의 계측과 물체 식별 능력을 높이기 위해서는 반도체 레이저의 출력을 높이고 펄스폭을 줄여서 분해능을 향상하는 것이 필요하다. 또한, 물체 인식 또는 계측 시간을 빠르게 하기 위해서는 고속 주파수로 동작하게 되면 가능해 질 것이다. 본 논문은 1,550 nm 대역의 반도체 레이저를 제작하여 펄스 구동으로 출력과 펄스폭을 측정하였다. 또한, 보다 높은 전류에서 칩 단면의 열화를 방지하기 위한 기술을 적용하여 설계 및 제작된 레이저의 특성을 측정하여 향후, 지능형 자동차의 레이저 레이다(LIDAR)와 같은 응용분야에 많이 활용될 수 있을 것으로 기대한다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.5 no.4
• /
• pp.308-316
• /
• 2004

As we know that there are so many restrictions controlling the autonomous robot to turn and move in an indoor space. In this research, Ive adopted the concept ‘Omni-directional wheel’ as a driving equipment, which makes it possible for the robot to move in horizontal and diagonal directions. Most of all, we eliminated the slip error problem, which can occur when the system generates power by means of slip. In order to solve this problem, we developed a ‘slip error correction algorithm’. Following this program, whenever the robot moves in any directions, it defines its course by comparing pre-programmed direction and the current moving way, which can be decided by extracted image of floor line. Additionally, this robot also provides the limited security and service function. It detects the motion of vehicle, transmits pictures to multiple users and can be moved by simple order's. In this paper, we tried to propose a practical model which can be used in an office.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.6
• /
• pp.1195-1200
• /
• 2007

This paper describes indoor location estimation intelligent robot. Indoor location estimation function using RSSI based indoor location estimation system and wireless sensor networks were implemented in the robot. Spartan III(Xilinx, U.S.A.) was used as a main control device in the mobile robot and the current direction data was collected in the indoor location estimation system. The data was transferred to the wireless sensor network node attached to the mobile robot through Zigbee/IEEE 802.15.4, a wireless communication. After receiving it, with the data of magnetic compass the node is aware of and senses the direction the robot head for and the robot moves to its destination. Indoor location estimation intelligent robot is can be moved efficiently and actively without obstacle on flat ground to the appointment position by user.

• Journal of Convergence for Information Technology
• /
• v.11 no.10
• /
• pp.45-52
• /
• 2021

In this paper, we proposed an Atrous Residual U-Net (AR-UNet) to improve the segmentation accuracy of semantic segmentation method based on U-Net. The U-Net is mainly used in fields such as medical image analysis, autonomous vehicles, and remote sensing images. The conventional U-Net lacks extracted features due to the small number of convolution layers in the encoder part. The extracted features are essential for classifying object categories, and if they are insufficient, it causes a problem of lowering the segmentation accuracy. Therefore, to improve this problem, we proposed the AR-UNet using residual learning and ASPP in the encoder. Residual learning improves feature extraction ability and is effective in preventing feature loss and vanishing gradient problems caused by continuous convolutions. In addition, ASPP enables additional feature extraction without reducing the resolution of the feature map. Experiments verified the effectiveness of the AR-UNet with Cityscapes dataset. The experimental results showed that the AR-UNet showed improved segmentation results compared to the conventional U-Net. In this way, AR-UNet can contribute to the advancement of many applications where accuracy is important.