Improved YOLOv8n for Gated Imaging Object Detection Algorithm
Laser gated imaging technology performs well in complex environments.But gated images cannot provide color information for grayscale images and have low contrast,so it is more difficult to detect small targets and occluded targets.In order to solve the above problems,an improved YOLOv8n gated image target detection algorithm is proposed.Firstly,in the backbone network part of feature extraction,the large kernel convolution C2f-DSF is used to capture the global information of the input data more effectively.Then,the multi-head attention detection head Detect-SEAM module is added to enhance the ability of feature extraction and object recognition.In addition,in order to obtain the context infor-mation of different receptive fields and enhance the feature extraction ability,the SPPF-M module is used.The upsam-pling operator Dysample is used to reduce the loss of feature information,so as to improve the detection accuracy of small objects.The improved YOLOv8n algorithm improves mAP@0.5 by 2.4 percentage points and mAP@0.5:0.95 by 1.8 per-centage points on the strobe image dataset.In order to verify the generalization of the improved YOLOv8n algorithm,the KITTI data set experiment is selected.Compared with the YOLOv8n algorithm,the improved YOLOv8n's mAP@0.5 is increased by 4.3 percentage points,and mAP@0.5:0.95 is increased by 3.5 percentage points.
万方数据
维普
