In the task of detecting targets under airport pavement， B-scan maps generated by Ground Penetrating Radar （GPR） have complex backgrounds and lots of noise， especially a single B-scan map cannot reflect the complete information of an underground target. To solve these problems， a Three-Dimensional Channel and Spatial Attention UNet （3D-CSA-UNet） model was established to automatically detect the underground targets. Firstly， a Three-Dimensional Channel and Spatial parallel attention Block （3D-CS-Block） was designed to make the model focus on the underground target information in radar C-scan and suppress the interference of backgrounds and noise. Secondly， in order to enhance the capability of 3D-CS-Block in feature extraction， a multi-scale 3D segmentation model was designed to extract feature maps of different sizes from the radar C-scan. Finally， the cross-entropy loss function was employed to calculate the loss value of feature map under each scale to improve the detection accuracy of the model. On a real dataset of targets under airport pavement， compared with 3D-Fully Convolutional Network （3D-FCN）， 3D-UNet and other algorithms， 3D-CSA-UNet has the average F1 score in terms of the pixel level segmentation for void， rebar and parallel rebar targets increased by at last 12.33， 9.05 and 11.05 percentage points. Experimental results show that 3D-CSA-UNet can meet the real engineering requirements well.

When using the Bscan image generated by Ground Penetrating Radar （GPR） to detect underground targets， the current target detection network models based on deep learning have some problems， such as high demand of training samples， long time consuming， unable to distinguish the significance of targets， and difficult to identify complex targets. To solve the above problems， a double threshold segmentation algorithm based on histogram was proposed. Firstly， based on the distribution characteristics of GPR image histogram of underground target， two thresholds for underground target segmentation were calculated quickly from the histogram. Then， a combination classifier model with Support Vector Machine （SVM） and LeNet was used to classify the segmentation results. Finally， classification results were integrated and the accuracy values were counted. Compared with the traditional threshold segmentation algorithms such as Ostu and iterative methods， the structure of the underground target segmentation results obtained by the proposed algorithm was more complete and almost free of noise. On the real dataset， the average recognition accuracy of the proposed algorithm reached more than 90%， which was more than 40% higher than that of the algorithm using a single classifier. The experimental results show that the salient and non-salient underground targets can be effectively segmented at the same time， and the combination classifier can obtain better classification results. It is suitable for automatic detection and recognition of underground targets with small sample datasets.