http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
( Alvaro Fuentes ),( Sook Yoon ),( Dong Sun Park ) 한국인터넷정보학회 2018 KSII Transactions on Internet and Information Syst Vol.12 No.12
Pedestrian detection is a challenging area in the intelligent vehicles domain. During the last years, many works have been proposed to efficiently detect motion in images. However, the problem becomes more complex when it comes to detecting moving areas while the vehicle is also moving. This paper presents a variational optical flow-based method for motion estimation in vehicular traffic scenarios. We introduce a framework for detecting motion areas with small and large displacements by computing optical flow using a multilevel architecture. The flow field is estimated at the shortest level and then successively computed until the largest level. We include a filtering parameter and a warping process using bicubic interpolation to combine the intermediate flow fields computed at each level during optimization to gain better performance. Furthermore, we find that by including a penalization function, our system is able to effectively reduce the presence of outliers and deal with all expected circumstances in real scenes. Experimental results are performed on various image sequences from Daimler Pedestrian Dataset that includes urban traffic scenarios. Our evaluation demonstrates that despite the complexity of the evaluated scenes, the motion areas with both moving and static camera can be effectively identified.
Data augmentation for plant growth prediction in time-series
Meng Yao,Alvaro Fuentes,Jaehwan Lee,Sook Yoon,Dong Sun Park 제어로봇시스템학회 2021 제어로봇시스템학회 각 지부별 자료집 Vol.2021 No.12
Deep learning technuques have been constantly developed and improved through the recent years. Many domains adopt these techniques for classification, object detection, segmentation, and prediction. Recently, agricultural industries utilize those techniques to improve plant harvesting and bring automation to farmers. One of the important tasks is plant growth prediction. Collecting labeled data from crops and greenhouses is a time-consuming and complicated task due to the large variety of crops, especially, when tasks require a large amount of labeled data for training. Data augmentation for training deep neural networks is a well-established technique. This paper presents a data augmentation method that holds the plant growth structure especially in time series, thus maintaining the plant physical appearance of the dataset as close as possible to plants in real agricultural scenes. The proposed method can be used in tasks that require plant image data in time series for plant growth prediction.
Efficient detection with attention mechanism for paprika disease diagnose
Jiuqing Dong,Alvaro Fuentes,Jaehwan Lee,Sook Yoon,Dong Sun Park 제어로봇시스템학회 2021 제어로봇시스템학회 각 지부별 자료집 Vol.2021 No.12
Deep learning technology is widely used in computer vision, especially in interdisciplinary fields, such as medicine, automatic driving and object detection, etc. In recent years, the combination of deep learning and agriculture has received more and more attention. In this work, we use a dataset of 6003 samples that covers five common diseases of paprika fruits and leaves. We applied Efficient-Det as the baseline detector, evaluated on this dataset. By using data enhancement, transfer learning, and attention mechanisms, the performance of the model can be greatly improved. The results indicated that our model is useful for paprika disease detection.
Multi-Cattle Tracking Algorithm with Enhanced Trajectory Estimation in Precision Livestock Farms
Shujie Han,Alvaro Fuentes,Sook Yoon,Jongbin Park,Dong Sun Park (사)한국스마트미디어학회 2024 스마트미디어저널 Vol.13 No.2
In precision cattle farm, reliably tracking the identity of each cattle is necessary. Effective tracking of cattle within farm environments presents a unique challenge, particularly with the need to minimize the occurrence of excessive tracking trajectories. To address this, we introduce a trajectory playback decision tree algorithm that reevaluates and cleans tracking results based on spatio-temporal relationships among trajectories. This approach considers trajectory as metadata, resulting in more realistic and accurate tracking outcomes. This algorithm showcases its robustness and capability through extensive comparisons with popular tracking models, consistently demonstrating the promotion of performance across various evaluation metrics that is HOTA, AssA, and IDF1 achieve 68.81%, 79.31%, and 84.81%.
Non-linear data association method for cow tracking
Shujie Han,Alvaro Fuentes,Jongbin Park,Sook Yoon,Dong Sun Park 제어로봇시스템학회 2021 제어로봇시스템학회 각 지부별 자료집 Vol.2021 No.12
Multi-objects tracking (MOT) is an important basic research area in computer vision in recent years. In particular, pedestrian tracking and vehicle tracking are very popular applications in MOT. In this paper, we perform MOT on cows, In this MOT task we address the following challenges: deformation and frequent occlusion. To solve deformation, we proposed a deep CNN model with SPP-Net as backbone to extract a feature vector of every cow. To solve occlusion, we use the ensemble Kalman filter that suitable for non-linear motion model to predict the attributes of bounding boxes directly. Our experiments reveal that nonlinear motion model and fixed deep features are more suitable on cows dataset with video sequences.
Intelligent Water and Nutrient Supply for Tomato Plant in Greenhouse
Mingle Xu,Alvaro Fuentes,Jongbin Park,Sook Yoon,Dong Sun Park 제어로봇시스템학회 2021 제어로봇시스템학회 각 지부별 자료집 Vol.2021 No.12
One of the challenges in the agricultural field is the method to supply nutrition and water. Excessive water and nutrition result in waste, high cost, and harm to our environment. The deficiency of them degrades the quality and yield of the plant and reduces the income of farmers. Currently, the water and nutrition supply is fixed in the greenhouse, instead of intelligent as the plant environment is not taken into consideration. In this paper, we aim to learn the relationship between the plant environment and the water nutrition supply curve deployed in current farms, which can be utilized to reduce drainage in the future. To achieve the goal, we leverage a long short-term memory network given time-series data from multimodal sensors. The experimental results show that the goal is achieved by our method
Multi-Cattle tracking with appearance and motion models in closed barns using deep learning
Shujie Han,박동선,Alvaro Fuentes,윤숙,박종빈 (사)한국스마트미디어학회 2022 스마트미디어저널 Vol.11 No.8
Precision livestock monitoring promises greater management efficiency for farmers and higher welfare standards for animals. Recent studies on video-based animal activity recognition and tracking have shown promising solutions for understanding animal behavior. To achieve that, surveillance cameras are installed diagonally above the barn in a typical cattle farm setup to monitor animals constantly. Under these circumstances, tracking individuals requires addressing challenges such as occlusion and visual appearance, which are the main reasons for track breakage and increased misidentification of animals. This paper presents a framework for multi-cattle tracking in closed barns with appearance and motion models. To overcome the above challenges, we modify the DeepSORT algorithm to achieve higher tracking accuracy by three contributions. First, we reduce the weight of appearance information. Second, we use an Ensemble Kalman Filter to predict the random motion information of cattle. Third, we propose a supplementary matching algorithm that compares the absolute cattle position in the barn to reassign lost tracks. The main idea of the matching algorithm assumes that the number of cattle is fixed in the barn, so the edge of the barn is where new trajectories are most likely to emerge. Experimental results are performed on our dataset collected on two cattle farms. Our algorithm achieves 70.37%, 77.39%, and 81.74% performance on HOTA, AssA, and IDF1, representing an improvement of 1.53%, 4.17%, and 0.96%, respectively, compared to the original method.
Towards Improved Performance on Plant Disease Recognition with Symptoms Specific Annotation
Jiuqing Dong,박동선,윤숙,Alvaro Fuentes,김태현 (사)한국스마트미디어학회 2022 스마트미디어저널 Vol.11 No.4
Object detection models have become the current tool of choice for plant disease detection in precision agriculture. Most existing research improves the performance by ameliorating networks and optimizing the loss function. However, the data-centric part of a whole project also needs more investigation. In this paper, we proposed a systematic strategy with three different annotation methods for plant disease detection: local, semi-global, and global label. Experimental results on our paprika disease dataset show that a single class annotation with semi-global boxes may improve accuracy. In addition, we also studied the noise factor during the labeling process. An ablation study shows that annotation noise within 10% is acceptable for keeping good performance. Overall, this data-centric numerical analysis helps us to understand the significance of annotation methods, which provides practitioners a way to obtain higher performance and reduce annotation costs on plant disease detection tasks. Our work encourages researchers to pay more attention to label quality and the essential issues of labeling methods.