Object-level SLAM introduces semantic meaningful and compact object landmarks that help both indoor robot applications and outdoor autonomous driving tasks. However, the back end of object-level SLAM suffers from singularity problems because existing methods parameterize object landmark separately by their scales and poses. Under that parameterization method, the same abstract object can be represented by rotating the object coordinate frame by 90 deg and swapping its length with width value, making the pose of the same object landmark not globally consistent. To avoid the singularity problem, we first introduce the symmetric positive-definite (SPD) matrix manifold as an improved object-level landmark representation and further improve the cost functions in the back end to make them compatible with the representation. Our method demonstrates a faster convergence rate and more robustness in simulation experiments. Experiments on real datasets also reveal that using the same front-end data, our strategy improves the mapping accuracy by 22% on average.