Crafting an edge-based real-time object detector for unmanned aerial vehicle (UAV) aerial images is challenging because of the limited computational resources and the small size of detected objects. Existing lightweight object detectors often prioritize speed over detecting extremely small targets. To better balance this trade-off, this paper proposes an efficient and low-complexity object detector for edge computing platforms deployed on UAVs, termed EUAVDet (Edge-based UAV Object Detector). Specifically, an efficient feature downsampling module and a novel multi-kernel aggregation block are first introduced into the backbone network to retain more feature details and capture richer spatial information. Subsequently, an improved feature pyramid network with a faster ghost module is incorporated into the neck network to fuse multi-scale features with fewer parameters. Experimental evaluations on the VisDrone, SeaDronesSeeV2, and UAVDT datasets demonstrate the effectiveness and plug-and-play capability of our proposed modules. Compared with the state-of-the-art YOLOv8 detector, the proposed EUAVDet achieves better performance in nearly all the metrics, including parameters, FLOPs, mAP, and FPS. The smallest version of EUAVDet (EUAVDet-n) contains only 1.34 M parameters and achieves over 20 fps on the Jetson Nano. Our algorithm strikes a better balance between detection accuracy and inference speed, making it suitable for edge-based UAV applications.

Multi-component cutting force sensing systems in manufacturing processes applied to cutting tools are gradually becoming the most significant monitoring indicator. Their signals have been extensively applied to evaluate the machinability of workpiece materials, predict cutter breakage, estimate cutting tool wear, control machine tool chatter, determine stable machining parameters, and improve surface finish. Robust and effective sensing systems with capability of monitoring the cutting force in machine operations in real time are crucial for realizing the full potential of cutting capabilities of computer numerically controlled (CNC) tools. The main objective of this paper is to present a brief review of the existing achievements in the field of multi-component cutting force sensing systems in modern manufacturing.

Sports analysis has recently attracted increasing research efforts in computer vision. Among them, basketball video analysis is very challenging due to severe occlusions and fast motions. As a typical tracking-by-detection method, k-shortest paths (KSP) tracking framework has been well used for multiple-person tracking. While effective and fast, the neglect of the appearance model would easily lead to identity switches, especially when two or more players are intertwined with each other. This paper addresses this problem by taking the appearance features into account based on the KSP framework. Furthermore, we also introduce a similarity measurement method that can fuse multiple appearance features together. In this paper, we select jersey color and jersey number as two example features. Experiments indicate that about 70% of jersey color and 50% of jersey number over a whole sequence would ensure our proposed method preserve the player identity better than the existing KSP tracking method.

The automatic reading recognition of pointer meters plays a crucial role in data monitoring and analysis in intelligent substations. Existing meter reading methods struggle to address challenging difficulties such as image distortion and varying illumination. To enhance their robustness and accuracy, this study proposes a novel approach that leverages the TransUNet semantic segmentation model and a perspective transformation correction method. Initially, the dial of the pointer meter is localized from the natural background using YOLOv8. Subsequently, after enhancing the image with Gamma correction technology, the scale lines and the pointer within the dial are extracted using the TransUNet model. The distorted or rotated dial can then be corrected through perspective transformation. Finally, the meter readings are accurately obtained by the Weighted Angle Method (WAM). Ablative and comparative experiments on two self-collected datasets clearly verify the effectiveness of the proposed method, with a reading accuracy of 97.81% on Simple-MeterData and 93.39% on Complex-MeterData, respectively.

Recognition and early warning of plant diseases is one of the keys to agricultural disaster prevention and mitigation. Deep learning-based image recognition methods give us a new idea for plant disease identification. Due to the harsh conditions in agricultural environment, recent research has focused on exploring ways to lightweight the recognition model for deployment on low-power devices. In this paper, we propose an efficient and feature-guided real-time plant disease recognition model with a multi-classifier architecture, specifically designed for low-power devices. By comparing with other advanced methods, our model reaches the state-of-the-art in the combined metrics of recognition accuracy, the number of parameters and inference speed. First, we propose AMI-NanoNet based on RoofLine theory to significantly reduce the number of parameters and computational complexity. This model can achieve 99.8343% accuracy on PlantVillage by using a Feature-Guided Curriculum Learning (FGCL) with stepwise training strategy. Moreover, we design another training strategy suitable for lightweight ensemble models. Based on this strategy, our model only needs to integrate the classifiers at the end of the network to achieve 99.8708% identification accuracy, and it hardly increases the number of operations and parameters of the network. Extensive evaluations on this dataset demonstrate the effectiveness of our ensemble learning method. Furthermore, we tested our proposed methods on another dataset from other domains to validate its applicability to different scenarios. Overall, our research provides a basis for rapid and intelligent identification of plant diseases.

Pansharpening is a spatial-spectral fusion technique that fuses low-resolution multispectral (MS) images with high-resolution panchromatic (PAN) images to get high-resolution multispectral images which are rich in spectral and spatial information. Some pansharpening methods based on dynamic convolution were proposed to improve the adaptivity and generalization of fusion network. However, these methods either only focus on local small regions or generate dynamic filters with a complex network. Besides, the dynamic filters in the existing methods only convolve with MS or PAN image, resulting in that the extracted details or spectral features are inadequate. In this article, we propose a dynamic large-small kernel convolutional network. To obtain small scale features, we propose a dual dynamic small kernel (DDSK) module which consists of dynamic spatial small filter (DASF) and dynamic spectral small filter (DESF). The multiscale dynamic large kernel convolution (MDLC) module is designed to expand the receptive field for obtaining large scale features. Considering the differences between PAN and MS images, DASF and spatial MDLC modules are presented to extract the details of PAN image. Similarly, DESF and spectral MDLC modules are proposed to obtain the spectral features of MS images. The proposed method is evaluated on GaoFen-2 and WorldView-3 datasets, and our method shows good performance.