• Journal of Internet of Things and Convergence
• /
• v.10 no.5
• /
• pp.235-252
• /
• 2024

To address the inefficiencies of visual interpretation for mining equipment supervision in open-pit mines, this study proposes an enhanced YOLOv8n+DeepSORT framework. By leveraging high-point surveillance for image acquisition and optimizing both YOLOv8n for equipment identification and DeepSORT for real-time tracking, we overcome limitations in accuracy, cost, and real-time monitoring. Field validation at Jingkai Runding Mine, Pingxiang, Jiangxi, demonstrates the technology's efficacy in identifying and tracking mining equipment, featuring rapid algorithm convergence, low computational overhead, and near-precise target detection and tracking. This approach paves the way for algorithmic support to facilitate effective government regulation of open-pit mining operations.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.10
• /
• pp.137-148
• /
• 2024

The European honeybee (Apis mellifera) is an important pollinator threatened by colony collapse disorder (CCD), primarily due to infestation by the Varroa mite (Varroa destructor). Traditional detection methods are invasive and time-consuming, often causing additional stress to colonies. We propose a two-stage framework using the You Only Look Once version 8 (YOLOv8) model for non-destructive and rapid detection of Varroa mite infestation. The framework uses comb light images from inside the hives. In the first stage, a YOLOv8-n model detects bees and extracts individual bee images. In the second stage, a YOLOv8-cls model classifies the infestation status of each bee. Our object detection model achieved a mAP@0.5 of 0.701, and the classification model achieved an average accuracy of 91%. These results demonstrate the effectiveness of the framework as a non-destructive method for Varroa mite detection. Based on this research, we expect to provide beekeepers with an efficient tool for early detection and management of Varroa mite infestations, potentially reducing the incidence of CCD and supporting the sustainability of apiculture.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.85-92
• /
• 2024

In this paper presents a comparative analysis of the performance between the latest object detection model, YOLOv10, and its previous versions. YOLOv10 introduces NMS-Free training, an enhanced model architecture, and an efficiency-centric design, resulting in outstanding performance. Experimental results using the COCO dataset demonstrate that YOLOv10-N maintains high accuracy of 39.5% and low latency of 1.84ms, despite having only 2.3M parameters and 6.7G floating-point operations (FLOPs). The key performance metrics used include the number of model parameters, FLOPs, average precision (AP), and latency. The analysis confirms the effectiveness of YOLOv10 as a real-time object detection model across various applications. Future research directions include testing on diverse datasets, further model optimization, and expanding application scenarios. These efforts aim to further enhance YOLOv10's versatility and efficiency.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.10
• /
• pp.2966-2983
• /
• 2024

Considering the intricate and ever-changing nature of the marine environment and the diverse range of sizes for targets involved in marine ship target recognition, which present challenges in detecting specific targets, a marine ship target detection algorithm has been developed based on an enhanced iteration of YOLOv5. Initially, the integration of dynamic snake convolution (DySnakeConv) into the feature extraction network and subsequent enhancement of the C3 module based on this integration were implemented. This integration enables dynamic adjustments based on the input image size, adaptive fusion of feature sequences, and resolution of accuracy and continuity issues during the recognition process. Subsequently, a novel hybrid encoder (FSI) was devised, utilizing target scale characteristics to enhance the extraction capability of multi-scale information, facilitating effective detection and recognition of objects within images. Finally, we selected the Shape-IOU bounding box loss function to mitigate fixed target frame issues and enhance target detection accuracy. Experimental evaluations were conducted utilizing the Infrared Maritime Ship dataset. The results demonstrated that our enhanced model achieved a prediction accuracy of 93.8% and an average precision (mAP) value of 93.89%, surpassing YOLOv8s by 1.2% and 1.8%, respectively. Moreover, there was an increase in recall rate by 2% compared to YOLOv8n while reducing parameters from 10,473,392 to 6,549,901 only. The computational load decreased by 6.3 GFLOps compared with YOLOV8n, resulting in better performance in ocean target detection and recognition.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.1035-1036
• /
• 2023

본 논문은 저시력자를 위한 체외진단 의료기기 결과 판독 시스템을 제안한다. 이 시스템은 YOLOv8n 객체 탐지 모델을 기반으로 하며, 라즈베리파이4B+에서 홈 디바이스 형태로 구현하였다. 사용자는 음성 및 물리 버튼을 통해 명령을 입력하고, 동작 감지를 통해 자동으로 체외진단 의료기기를 촬영하여 학습된 모델로 결과를 판독하고 해당 결과를 사용자에게 출력한다. 또한, 판독 결과물과 함께 검사 일시 및 의료기기 종류를 데이터베이스에 저장하여 사용자에게 보다 높은 편의성을 제공한다.