The Aeroplane and Undercarriage Detection Based on Attention Mechanism and Multi-Scale Features Processing

Gao, Ruizhen; Zhang, Shuai; Wang, Haoqian; Zhang, Jingjun; Li, Hui; Zhang, Zhongqi

doi:10.1155/2022/2582288

Cited by 3 publications

(2 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, it can meet the real-time requirements of the detection system. We compared the data of the improved model with the undercarriage detection model proposed by Gao et al [ 42 ]. It is evident from the experimental results that our proposed model surpasses Gao’s model in terms of AP, mAP, and FPS metrics, demonstrating the effectiveness of YOLOv5-RSC.…”

Section: Experiments and Resultsmentioning

confidence: 99%

Real-Time Detection of an Undercarriage Based on Receptive Field Blocks and Coordinate Attention

Gao,

Ma,

Zhao

et al. 2023

Sensors

View full text Add to dashboard Cite

Currently, aeroplane images captured by camera sensors are characterized by their small size and intricate backgrounds, posing a challenge for existing deep learning algorithms in effectively detecting small targets. This paper incorporates the RFBNet (a coordinate attention mechanism) and the SIOU loss function into the YOLOv5 algorithm to address this issue. The result is developing the model for aeroplane and undercarriage detection. The primary goal is to synergize camera sensors with deep learning algorithms, improving image capture precision. YOLOv5-RSC enhances three aspects: firstly, it introduces the receptive field block based on the backbone network, increasing the size of the receptive field of the feature map, enhancing the connection between shallow and deep feature maps, and further improving the model’s utilization of feature information. Secondly, the coordinate attention mechanism is added to the feature fusion network to assist the model in more accurately locating the targets of interest, considering attention in the channel and spatial dimensions. This enhances the model’s attention to key information and improves detection precision. Finally, the SIoU bounding box loss function is adopted to address the issue of IoU’s insensitivity to scale and increase the speed of model bounding box convergence. Subsequently, the Basler camera experimental platform was constructed for experimental verification. The results demonstrate that the AP values of the YOLOv5-RSC detection model for aeroplane and undercarriage are 92.4% and 80.5%, respectively. The mAP value is 86.4%, which is 2.0%, 5.4%, and 3.7% higher than the original YOLOv5 algorithm, respectively, with a detection speed reaching 89.2 FPS. These findings indicate that the model exhibits high detection precision and speed, providing a valuable reference for aeroplane undercarriage detection.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Real-Time Detection of an Undercarriage Based on Receptive Field Blocks and Coordinate Attention

Gao,

Ma,

Zhao

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…The availability of a large number of remote sensing images (RSI) promotes correlation study in comprehending the content of remote sensing images, including scene classification, image retrieval [1], aeroplane recognition [2], detection of vehicles [3], recognition of building [4] etc. In these applications, objectrelated data is evaluated from very high resolution (VHR) RSI, which requires object detection [5], which is both fundamental and important.…”

Section: Introductionmentioning

confidence: 99%

Performance Analysis of Different Optimization Algorithms for Multi-Class Object Detection

Borade

Muddana

2023

IJRITCC

View full text Add to dashboard Cite

Object recognition is a significant approach employed for recognizing suitable objects from the image. Various improvements, particularly in computer vision, are probable to diagnose highly difficult tasks with the assistance of local feature detection methodologies. Detecting multi-class objects is quite challenging, and many existing researches have worked to enhance the overall accuracy. But because of certain limitations like higher network loss, degraded training ability, improper consideration of features, less convergent and so on. The proposed research introduced a hybrid convolutional neural network (H-CNN) approach to overcome these drawbacks. The collected input images are pre-processed initially through Gaussian filtering to eradicate the noise and enhance the image quality. Followed by image pre-processing, the objects present in the images are localized using Grid Guided Localization (GGL). The effective features are extracted from the localized objects using the AlexNet model. Different objects are classified by replacing the concluding softmax layer of AlexNet with Support Vector Regression (SVR) model. The losses present in the network model are optimized using the Improved Grey Wolf (IGW) optimization procedure. The performances of the proposed model are analyzed using PYTHON. Various datasets are employed, including MIT-67, PASCAL VOC2010, Microsoft (MS)-COCO and MSRC. The performances are analyzed by varying the loss optimization algorithms like improved Particle Swarm Optimization (IPSO), improved Genetic Algorithm (IGA), and improved dragon fly algorithm (IDFA), improved simulated annealing algorithm (ISAA) and improved bacterial foraging algorithm (IBFA), to choose the best algorithm. The proposed accuracy outcomes are attained as PASCAL VOC2010 (95.04%), MIT-67 dataset (96.02%), MSRC (97.37%), and MS COCO (94.53%), respectively.

show abstract

Real-time Detection Algorithm of Aircraft Landing Gear based on improved YOLOv8

Gao,

Chen,

Zhao

et al. 2024

Preprint

View full text Add to dashboard Cite

This study innovatively improves the YOLOv8 target detection model, aiming to achieve fast and accurate detection of aircraft landing gear in natural environments. By introducing a small target detection layer, a dynamic serpentine convolutional layer, and a CoTattention mechanism, the study successfully optimized the original yolov8 model to effectively detect small-sized aircraft landing gears when presented at a distance. This paper introduces a small target detection layer of 160x160 on top of the original network, significantly improving the detection performance of airplane landing gear by fusing features from different layers. Dynamic serpentine convolution uses a dynamic structure and iterative strategy to improve the model's ability to perceive complex geometric structures by optimizing the convolution kernel. The CoTAttention mechanism allows the model to consider the information of each position in the input image more comprehensively. It significantly reduces the loss of contextual information by enhancing the ability to perceive small targets. The experimental findings demonstrate a noteworthy enhancement in the performance metrics, including precision, recall, and average accuracy, when comparing the enhanced model to its original counterpart. Furthermore, the improved model effectively meets the real-time detection requirements. Compared to other object detection models, the improved model performs, offering high accuracy and real-time detection capabilities, particularly demonstrating its versatility and practical value in detecting aircraft landing gear.

show abstract

The Aeroplane and Undercarriage Detection Based on Attention Mechanism and Multi-Scale Features Processing

Cited by 3 publications

References 28 publications

Real-Time Detection of an Undercarriage Based on Receptive Field Blocks and Coordinate Attention

Real-Time Detection of an Undercarriage Based on Receptive Field Blocks and Coordinate Attention

Performance Analysis of Different Optimization Algorithms for Multi-Class Object Detection

Real-time Detection Algorithm of Aircraft Landing Gear based on improved YOLOv8

Contact Info

Product

Resources

About