提出了一种提高YOLO-v8雷达图像中小目标检测性能的创新方法.首先,对原始图像应用局部直方图均衡化技术,显著增强对比度和细节表示.然后,通过结合基于多维注意力机制和并行处理策略的卷积核增强YOLO-v8骨干网络实现更有效的特征信息融合.在模型上部添加了一个上采样层,与浅层网络输出融合,设计一个专门为小物体检测量身定制的检测头,从而进一步提高了精度.此外,对损失函数进行了修改,将局部交并比(Intersection over union,IoU)与尺度IoU结合使用,从而提高了模型的性能.引入了加权策略,有效地提高了小目标的检测精度.实验结果表明,定制模型在各种评估指标上优于传统方法,包括召回率、精确度、F1评分和受试者特征(Receiver operating characteristic,ROC)曲线,验证了其在雷达图像中小目标检测方面的有效性和创新性.结果表明,与图像分割和标准卷积神经网络等传统方法相比,所提方法准确性有了显著提高.
YOLO-v8 with Multidimensional Attention and Upsampling Fusion for Small Air Target Detection in Radar Images
This study presents an innovative approach to improving the performance of YOLO-v8 model for small object detection in radar images.Initially,a local histogram equalization technique was applied to the original images,resulting in a notable enhancement in both contrast and detail representation.Subsequently,the YOLO-v8 backbone network was augmented by incorporating convolutional kernels based on a multidimensional attention mechanism and a parallel processing strategy,which facilitated more effective feature information fusion.At the model's head,an upsampling layer was added,along with the fusion of outputs from the shallow network,and a detection head specifically tailored for small object detection,thereby further improving accuracy.Additionally,the loss function was modified to incorporate focal-intersection over union(IoU)in conjunction with scaled-IoU,which enhanced the model's performance.A weighting strategy was also introduced,effectively improving detection accuracy for small targets.Experimental results demonstrate that the customized model outperforms traditional approaches across various evaluation metrics,including recall,precision,F1-score,and the receiver operating characteristic(ROC)curve,validating its efficacy and innovation in small object detection within radar imagery.The results indicate a substantial improvement in accuracy compared to conventional methods such as image segmentation and standard convolutional neural networks.
国家科技期刊平台
NSTL
万方数据
