Vehicle Detection using Images taken by Low-Altitude Unmanned Aerial Vehicles (UAVs)

Seo, Chang Jin

doi:10.17485/ijst/2016/v9i24/96057

Cited by 3 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Object detection in low-altitude UAV images is a crucial problem because apart from the small representations, the heterogeneity of input images also makes the task more difficult [34]. For instance, an aerial image can be in different resolutions, which hinders the detector from detecting small objects.…”

Section: Challenges Of Low-altitude Uav Object Detectionmentioning

confidence: 99%

See 1 more Smart Citation

A Feature Pyramid Based Multi-stage Framework for Object Detection in Low-altitude UAV Images

Mittal

Sharma

Singh

2022

Int. J. Artif. Intell. Tools

View full text Add to dashboard Cite

This paper proposes a high-performance framework for accurate multi-stage object detection in low-altitude based UAV images. The proposed system employs a cascade style architecture with increasing thresholds for achieving accurate detection. The framework makes use of highly efficient Feature Pyramid Networks (FPNs) to detect objects of small sizes, and various scales which are the main challenge in low-altitude aerial images. FPNs aim to resolve scale variation problems in object detection by combining features of multiple levels. The experiments have been performed on a complex low-altitude aerial dataset VisDrone which has multiple categories of classes. The FPN-Cascade detector has been supported by slicing the data horizontally and vertically that resulted in an advancement of 8% mAP when compared with the base detector. The experiments compare the FPN-Cascade performance on the standard as well as augmented VisDrone dataset. A concrete methodology about the training process, hyperparameter tuning, and performance evaluation methods for Cascade RCNN on the VisDrone dataset is highlighted. The proposed framework achieves state of the art 30.04% mAP value on the VisDrone dataset.

show abstract

Section: Challenges Of Low-altitude Uav Object Detectionmentioning

confidence: 99%

“…VGG16 [32], ResNet50, ResNet101 [33], DarkNet53 [34] A backbone is pre-trained on standard dataset ImageNet Neck FPN [23], PANet [35], Bi-FPN [36] The neck are some layers between backbone and head to collect feature maps from different stages.…”

Section: Backbonementioning

confidence: 99%