Abstract: With the development of radar technology, a single radar can now simultaneously possess narrowband tracking and wideband measurement capabilities. How to fully utilize characteristic data, such as narrowband RCS and wideband HRRP, to target recognition has become a core issue that needs to be addressed in the design of recognition systems. In response to the technical challenge, this paper proposes a target recognition algorithm that integrates multiple feature data. The algorithm fully considers the different characteristics of RCS data and HRRP data. Although RCS data is easy to measure and abundant, the target information in it is relatively rough. In contrast, HRRP data is rich in information and has distinct features, but it suffers from data deficiency issue. Therefore, the research team innovatively adopts a hybrid recognition strategy, utilizing neural network algorithms with strong pattern recognition capabilities for fuzzy target recognition, and template matching algorithms for the probabilistic determination of target attributes, based on the characteristics of RCS data and HRRP data, respectively. Finally, the two recognition results are effectively fused through evidence theory, fully leveraging the complementary advantages of different data sources. Experimental results show that this system design not only significantly improves the accuracy of target recognition but also enhances the stability of the recognition process, providing a new technical approach for radar target recognition in complex battlefield environments.