In order to obtain the shape and dimension information of low dynamic non-resolved object in near space,an experiment on optical detection of high altitude balloon was performed to investigate how to calculate object shape and dimension information from photometric data. In the experiment, experimental images captured by the ground-based detector were processed by aperture photometry technology, thus obtaining non-resolved photometric data. During the inversion process, two shape representation methods as well as three sets of regularization functions were adopted to describe and restrict object shape, respectively. The methods contained spherical harmonics function and subdivision control points, and the functions involved spherical harmonics regularization function, triangle panel regularization function and physical characteristics regularization function. Based on the Fourier transform of photometric data and model data of the shape representation methods, the shape and dimension information of the non-resolved object was inversed by the Point Spread Function (PSF) of the optical system. The result shows that the prominent features presented in the object shape are similar, which indicates that the features are extracted from the photometric data. The relative errors of object equivalent diameter inversed by spherical harmonics function and subdivision control points are 11.3% and 22.6%, and the relative errors of object length are 11.6% and 21.8%, respectively. The comparison of the relative errors proves that the shape representation method based on the spherical harmonics function has smaller error and is more suitable to represent the low dynamic object shape. © 2017, Science Press. All right reserved.