The scattering coefficient of Lightning Whistler(LW)is a critical physical parameter for inverting space electron density. To achieve precise and automatic extraction of the scattering coefficient from massive observational data, this paper proposes an Automatic Extraction Algorithm for Lightning Whistler Scattering Coefficients(LWSC)utilizing computer vision techniques, which is applied to Zhangheng satellite data. Firstly, short-time Fourier transform processing is employed to obtain time-frequency maps. Then, a semantic segmentation network is used to detect the time-frequency trajectories of each LW. Subsequently, a quadratic polynomial regression analysis is performed on the time-frequency trajectory points to derive the scattering coefficients of each LW. Following this, satellite data from April 1st to April 5th, 2020, are utilized to evaluate the performance of the LWSC model, and experiments reveal that compared to manual extraction results, the mean absolute error is less than 0.0045 and the mean relative error is less than 0.0121.The average time of single piece detection is 0.98 s, which is better than the mainstream models in this field in terms of speed and accuracy.Finally, the algorithm is applied to Zhangheng satellite data from April 5th to April 10th, 2020, to automatically acquire the scattering coefficients of LWs over one orbital period. Analysis leads to the following conclusions: The overall distribution of LW scattering coefficients exhibits a bimodal pattern: the values of the scattering coefficients are higher near latitudes of ±20°, and they gradually decrease with increasing latitude, showing a similar trend to the distribution of electron density in the space ionosphere. This result further validates the effectiveness and practicality of the algorithm for probing space electron density, holding significant reference value and application potential for the automatic mining and analysis of physical parameters from vast amounts of space electromagnetic events.