Structural deformation， as one of the important indicators of structural health monitoring， is crucial for evaluating structural strength performance. In practical engineering， due to engineering use and other factors， the number and placement of strain sensors are limited， resulting in reduced accuracy of structural deformation reconstruction. To address the above problem， this paper proposes a strain field reconstruction algorithm based on two-node inverse beam elements. First， the structure is discretized using two-node beam elements. An error function is constructed based on the principles of Timoshenko beam theory and least squares error theory， which evaluates the discrepancy between the theoretical strains within the element and the measured strains. The strain field distribution of the structure is reconstructed using discrete strain measurement points， and this serves as the foundation for reconstructing the structural deformations. The effectiveness of the algorithm is validated through simulations in ABAQUS and experimental tests. The results indicate that the algorithm can effectively reconstruct structural deformations even with a limited number of strain measurement points， reducing the relative error by 5% compared with directly using the strain data from measurement points. It has significant implications for structural health monitoring.