The dynamic change of UAV flight height makes aerial images often contain a large number of small targets, and the target scale changes significantly. These problems bring challenges to the target detection task. In view of the above problems, this paper proposes a UAV aerial target detection method based on multi-scale fusion and high-resolution feature enhancement. Firstly, a multi-scale structure re-parameterized feature extraction module is introduced into the backbone network. The convolution operation of different scales is performed on multiple branches by using ordinary convolution blocks and structure re-parameterized large-core convolution blocks, and the feature information under different receptive fields is effectively extracted. Then, a multi-dimensional feature adaptive fusion module based on the feature pyramid network is introduced into the neck network to optimize its bottom-up feature aggregation process, so as to realize the adaptive selection of fine details in shallow features and context information in deep features, so as to deal with the significant change of target scale more effectively. Finally, a multi-scale feature fusion small target enhancement module is introduced in the neck network to capture the changes of small target objects in UAV aerial images at different scales. Extensive experiments on two public datasets, VisDrone2019 and TinyPerson, demonstrate the effectiveness and superiority of the proposed method.