• Journal of the Korea Society of Computer and Information
• /
• v.27 no.1
• /
• pp.1-8
• /
• 2022

In this paper, we propose a fire detection system based on CCTV images using an object tracking technology with YOLOv4 model capable of real-time object detection and a DeepSORT algorithm. The fire detection model was learned from 10800 pieces of learning data and verified through 1,000 separate test sets. Subsequently, the fire detection rate in a single image and fire detection maintenance performance in the image were increased by tracking the detected fire area through the DeepSORT algorithm. It is verified that a fire detection rate for one frame in video data or single image could be detected in real time within 0.1 second. In this paper, our AI fire detection system is more stable and faster than the existing fire accident detection system.

• The Journal of the Korea Contents Association
• /
• v.22 no.10
• /
• pp.1-10
• /
• 2022

Visually impaired people have very difficulty using retail stores due to the absence of braille information on products and any other support system. In this paper, we propose a basic algorithm for a system that recognizes products in retail stores and explains them as a voice. First, the deep learning model detects hand objects and product objects in the input image. Then, it finds a product object that most overlapping hand object by comparing the coordinate information of each detected object. We determine that this is a product selected by the user, and the system read the nutritional information of the product as Text-To-Speech. As a result of the evaluation, we confirmed a high performance of the learning model. The proposed algorithm can be actively used to build a system that supports the use of retail stores for the visually impaired.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.9
• /
• pp.1330-1339
• /
• 2022

3D object detection generally aims to detect relatively large data such as automobiles, buses, persons, furniture, etc, so it is vulnerable to small object detection. In addition, in an environment with limited resources such as embedded devices, it is difficult to apply the model because of the huge amount of computation. In this paper, the accuracy of small object detection was improved by focusing on local features using only one layer, and the inference speed was improved through the proposed knowledge distillation method from large pre-trained network to small network and adaptive quantization method according to the parameter size. The proposed model was evaluated using SUN RGB-D Val and self-made apple tree data set. Finally, it achieved the accuracy performance of 62.04% at mAP@0.25 and 47.1% at mAP@0.5, and the inference speed was 120.5 scenes per sec, showing a fast real-time processing speed.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.2
• /
• pp.121-129
• /
• 2023

In this paper, we developed a traffic signal control system for emergency situations that can minimize loss of property and life by actively controlling traffic signals in a certain section in response to emergency situations. When the emergency vehicle terminal transmits an emergency signal including identification information and GPS information, the surrounding image is obtained from the camera, and the object is analyzed based on deep learning to output object information having information such as the location, type, and size of the object. After generating information tracking this object and detecting the signal system, the signal system is switched to emergency mode to identify and track the emergency vehicle based on the received GPS information, and to transmit emergency control signals based on the emergency vehicle's traveling route. It is a system that can be transmitted to a signal controller. This system prevents the emergency vehicle from being blocked by an emergency control signal that is applied first according to an emergency signal, thereby minimizing loss of life and property due to traffic obstacles.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.12
• /
• pp.1794-1799
• /
• 2022

Recently, the increasing number of overloaded vehicles on the road poses a risk to traffic safety, such as falling objects, road damage, and chain collisions due to the abnormal weight distribution, and can cause great damage once an accident occurs. therefore we propose to build an object detection-based AI model to identify overloaded vehicles that cause such social problems. In addition, we present a simple yet effective method to construct an object detection model for the large-scale vehicle images. In particular, we utilize the large-scale of vehicle image sets provided by open AI-Hub, which include the overloaded vehicles. We inspected the specific features of sizes of vehicles and types of image sources, and pre-processed these images to train a deep learning-based object detection model. Also, we propose an integrated system for tracking the detected vehicles. Finally, we demonstrated that the detection performance of the overloaded vehicle was improved by about 23% compared to the one using raw data.

• Journal of KIBIM
• /
• v.13 no.4
• /
• pp.96-105
• /
• 2023

This study focuses on developing a building object recognition technology for efficient use in the remodeling of buildings constructed without drawings. In the era of the 4th industrial revolution, smart technologies are being developed. This research contributes to the architectural field by introducing a deep learning-based method for automatic object classification and recognition, utilizing point cloud data. We use a TD3D network with voxels, optimizing its performance through adjustments in voxel size and number of blocks. This technology enables the classification of building objects such as walls, floors, and roofs from 3D scanning data, labeling them in polygonal forms to minimize boundary ambiguities. However, challenges in object boundary classifications were observed. The model facilitates the automatic classification of non-building objects, thereby reducing manual effort in data matching processes. It also distinguishes between elements to be demolished or retained during remodeling. The study minimized data set loss space by labeling using the extremities of the x, y, and z coordinates. The research aims to enhance the efficiency of building object classification and improve the quality of architectural plans by reducing manpower and time during remodeling. The study aligns with its goal of developing an efficient classification technology. Future work can extend to creating classified objects using parametric tools with polygon-labeled datasets, offering meaningful numerical analysis for remodeling processes. Continued research in this direction is anticipated to significantly advance the efficiency of building remodeling techniques.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.2
• /
• pp.53-61
• /
• 2016

Extracting focused object with saliency map is still remaining as one of the most highly tasked research area around computer vision for it is hard to estimate. Through this paper, we propose enhanced object extraction method based on multi-channel saliency map which could be done automatically without machine learning. Proposed Method shows a higher accuracy than Itti method using SLIC, Euclidean, and LBP algorithm as for object extraction. Experiments result shows that our approach is possible to be used for automatic object extraction without any previous training procedure through focusing on the main object from the image instead of estimating the whole image from background to foreground.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.54-57
• /
• 2022

Korea has a vulnerable environment from the risk of wildfires, which causes great damage every year. To prevent this, a lot of manpower is being used, but the effect is insufficient. If wildfires are detected and extinguished early through artificial intelligence technology, damage to property and people can be prevented. In this paper, we studied the object double detection method with the goal of minimizing the data collection and processing process that occurs in the process of creating an object detection model to minimize the damage of wildfires. In YOLOv5, the original image is primarily detected through a single model trained on a limited image, and the object detected in the original image is cropped through Crop. The possibility of improving the false positive object detection rate was confirmed through the object double detection method that re-detects the cropped image.

• ETRI Journal
• /
• v.45 no.5
• /
• pp.847-861
• /
• 2023

Dynamic object detection is essential for ensuring safe and reliable autonomous driving. Recently, light detection and ranging (LiDAR)-based object detection has been introduced and shown excellent performance on various benchmarks. Although LiDAR sensors have excellent accuracy in estimating distance, they lack texture or color information and have a lower resolution than conventional cameras. In addition, performance degradation occurs when a LiDAR-based object detection model is applied to different driving environments or when sensors from different LiDAR manufacturers are utilized owing to the domain gap phenomenon. To address these issues, a sensor-fusion-based object detection and classification method is proposed. The proposed method operates in real time, making it suitable for integration into autonomous vehicles. It performs well on our custom dataset and on publicly available datasets, demonstrating its effectiveness in real-world road environments. In addition, we will make available a novel three-dimensional moving object detection dataset called ETRI 3D MOD.

• Journal of Broadcast Engineering
• /
• v.24 no.7
• /
• pp.1266-1275
• /
• 2019

With increasing criminal rates and number of CCTVs, much attention has been paid to intelligent surveillance system on the horizon. Object detection and tracking algorithms have been developed to reduce false alarms and accurately help security agents immediately response to undesirable changes in video clips such as crimes and accidents. Many studies have proposed a variety of algorithms to improve accuracy of detecting and tracking objects outside tunnels. The proposed methods might not work well in a tunnel because of low illuminance significantly susceptible to tail and warning lights of driving vehicles. The detection performance has rarely been tested against the tunnel environment. This study investigated a feasibility of object detection and tracking in an actual tunnel environment by utilizing YOLOv3 and Kernelized Correlation Filter. We tested 40 actual video clips to differentiate pedestrians and motorcycles to evaluate the performance of our algorithm. The experimental results showed significant difference in detection between pedestrians and motorcycles without false positive rates. Our findings are expected to provide a stepping stone of developing efficient detection algorithms suitable for tunnel environment and encouraging other researchers to glean reliable tracking data for smarter and safer City.