润色：To solve this problem first of all the encoder extracts more abundant global information by using CNN and Transformer modules Secondly using the jump connection between the encoder and the decoder

To address this issue, a two-step approach is proposed. Firstly, the encoder module is enhanced by incorporating both CNN and Transformer modules, allowing for the extraction of a wider range of global information. This ensures that the encoder captures more comprehensive features necessary for accurate segmentation.

Secondly, a jump connection is established between the encoder and decoder modules. This connection enables the sharing of high-level features, facilitating the integration of these features into the decoding process. As a result, the segmentation performance of the decoder is improved while minimizing information loss from the encoder.

By combining these enhancements, the proposed method achieves better segmentation results and more efficient information transfer between the encoder and decoder