Abstract
With great significance in military and civilian applications, the topic of detecting small and densely arranged objects in wide-scale remote sensing imagery is still challenging nowadays. To solve this problem, we propose a novel effectively optimized one-stage network (NEOON). As a fully convolutional network, NEOON consists of four parts: Feature extraction, feature fusion, feature enhancement, and multi-scale detection. To extract effective features, the first part has implemented bottom-up and top-down coherent processing by taking successive down-sampling and up-sampling operations in conjunction with residual modules. The second part consolidates high-level and low-level features by adopting concatenation operations with subsequent convolutional operations to explicitly yield strong feature representation and semantic information. The third part is implemented by constructing a receptive field enhancement (RFE) module and incorporating it into the fore part of the network where the information of small objects exists. The final part is achieved by four detectors with different sensitivities accessing the fused features, all four parallel, to enable the network to make full use of information of objects in different scales. Besides, the Focal Loss is set to enable the cross entropy for classification to solve the tough problem of class imbalance in one-stage methods. In addition, we introduce the Soft-NMS to preserve accurate bounding boxes in the post-processing stage especially for densely arranged objects. Note that the split and merge strategy and multi-scale training strategy are employed in training. Thorough experiments are performed on ACS datasets constructed by us and NWPU VHR-10 datasets to evaluate the performance of NEOON. Specifically, 4.77% and 5.50% improvements in mAP and recall, respectively, on the ACS dataset as compared to YOLOv3 powerfully prove that NEOON can effectually improve the detection accuracy of small objects in remote sensing imagery. In addition, extensive experiments and comprehensive evaluations on the NWPU VHR-10 dataset with 10 classes have illustrated the superiority of NEOON in the extraction of spatial information of high-resolution remote sensing images.
Highlights
Remote sensing imaging technology, such as optical or hyperspectral aerial image processing [1,2,3], has rapidly become one of the most significant technologies in image processing, especially in object detection [4]
These results demonstrate the high superiority of novel effectively optimized one-stage network (NEOON) achieving better performance compared to the existing widely used methods in remote sensing object detection
We proposed the NEOON, which is a novel one-stage model designed and optimized for extracting spatial information of high-resolution remote sensing images by understanding and analyzing the combination of feature and semantic information of small objects
Summary
Remote sensing imaging technology, such as optical or hyperspectral aerial image processing [1,2,3], has rapidly become one of the most significant technologies in image processing, especially in object detection [4]. With the recent advent of large ground-based datasets and advanced computational techniques, methods [16,17] based on deep neural network, especially convolutional neural network, (CNN) have achieved great success in general object detection. There are two main streams of CNN-based object detection methods: The two-stage frameworks and the one-stage frameworks. To accelerate the detection process, the one-stage frameworks, including You Only Look Once (YOLO) [22,23,24] and Single Shot Multi-Box Detector (SSD) [25], directly predict bounding boxes and produce detection results simultaneously. Compared to the two-stage frameworks, YOLO and SSD run faster but tend to sacrifice detection accuracy to a certain extent
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
