An Efficient Deep Learning Framework for People Detection in Overhead Images

Peker, Musa; İnci, Bilge; Musaoglu, Elnura; Çobanoğlu, Hüseyin; Kocakır, Nadir; Karademir, Önder

doi:10.1007/978-3-030-85383-9_1

Cited by 5 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In an effort to balance precision and computational efficiency for embedded applications, this research introduces the utilization of CBAM, which enhances feature extraction capabilities, particularly beneficial for identifying small objects and complex backgrounds. The model employs DIoU-NMS in lieu of the traditional NMS [49]. This approach considers the distance between bounding box centers, which not only retains targets more effectively but also distinguishes overlapping boxes with greater precision.…”

Section: Discussionmentioning

confidence: 99%

Research on Real-Time Detection of Maize Seedling Navigation Line Based on Improved YOLOv5s Lightweighting Technology

Gong,

Wang,

Zhuang

2024

Agriculture

View full text Add to dashboard Cite

This study focuses on real-time detection of maize crop rows using deep learning technology to meet the needs of autonomous navigation for weed removal during the maize seedling stage. Crop row recognition is affected by natural factors such as soil exposure, soil straw residue, mutual shading of plant leaves, and light conditions. To address this issue, the YOLOv5s network model is improved by replacing the backbone network with the improved MobileNetv3, establishing a combination network model YOLOv5-M3 and using the convolutional block attention module (CBAM) to enhance detection accuracy. Distance-IoU Non-Maximum Suppression (DIoU-NMS) is used to improve the identification degree of the occluded targets, and knowledge distillation is used to increase the recall rate and accuracy of the model. The improved YOLOv5s target detection model is applied to the recognition and positioning of maize seedlings, and the optimal target position for weeding is obtained by max-min optimization. Experimental results show that the YOLOv5-M3 network model achieves 92.2% mean average precision (mAP) for crop targets and the recognition speed is 39 frames per second (FPS). This method has the advantages of high detection accuracy, fast speed, and is light weight and has strong adaptability and anti-interference ability. It determines the relative position of maize seedlings and the weeding machine in real time, avoiding squeezing or damaging the seedlings.

show abstract

Section: Discussionmentioning

confidence: 99%

Research on Real-Time Detection of Maize Seedling Navigation Line Based on Improved YOLOv5s Lightweighting Technology

Gong,

Wang,

Zhuang

2024

Agriculture

View full text Add to dashboard Cite

show abstract

“…rough 3D laser scanning, we can directly obtain the image information in the murals, instead of traditional manual drawing, which will save a lot of manpower, material resources, and time [8]. is field has a great potential in application and development.…”

Section: Introductionmentioning

confidence: 99%

Application of Deep Learning Intelligent Laser Scanning Technology in Mural Digitization

Dong

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Ancient Chinese murals have a long history and a large number of types. They are witnesses to the development of ancient Chinese civilization. They still have important historical, artistic, and scientific values, as well as other values such as cultural relics, economy, and missionary education. Yet, the clear and smooth images of earlier have been damaged and covered by fog after the ancient murals were destroyed by human being and the nature. Therefore, the ancient murals that exist till now have been bothered by different damages and sickness. Protection and prevention are needed the most. In recent years, the application of scientific and technological means has played a role in the protection of ancient murals, making the work methods of cultural relics protection more scientific and diverse. In the context of the increasingly rich digital protection of cultural relics, the protection of murals requires more innovative work. However, at present, the resolution of ancient murals is low and the texture details are ambiguous, which leads to the problems of insufficient viewing and low research value. This paper focuses on ancient Chinese murals and conducts exploration on the phenomenon of mural damage and blurred colors. The deep learning intelligent laser scanning technology is used to extract the damaged ancient mural images. In this thesis, the images of murals have been restored using the new-super resolution technology to achieve the optimal mural images so that the mural images are more beautiful and artistic, which provides new ideas for the protection of murals.

show abstract

RescueNet: YOLO-based object detection model for detection and counting of flood survivors

Prabhu

Lakshmi

Ankitha

et al. 2022

Model. Earth Syst. Environ.

View full text Add to dashboard Cite

An Efficient Deep Learning Framework for People Detection in Overhead Images

Cited by 5 publications

References 27 publications

Research on Real-Time Detection of Maize Seedling Navigation Line Based on Improved YOLOv5s Lightweighting Technology

Research on Real-Time Detection of Maize Seedling Navigation Line Based on Improved YOLOv5s Lightweighting Technology

Application of Deep Learning Intelligent Laser Scanning Technology in Mural Digitization

RescueNet: YOLO-based object detection model for detection and counting of flood survivors

Contact Info

Product

Resources

About