In order to improve the positioning accuracy in the indoor environment of visible light communication (VLC), an improved location fingerprint position algorithm based on the combination of bilinear interpolation and K-means clustering algorithm is proposed. The algorithm first establishes the initial fingerprint, then interpolates the fingerprint in the cluster domain where the location point is located, and finally selects a matching algorithm to achieve the final location by taking the partial regional fingerprint after interpolation as the reference fingerprint inventory. The VLC system model is established in the indoor room of 5 m×5 m×3 m, and the influence of fingerprint density on positioning accuracy and the comparison of positioning accuracy of the improved algorithm are simulated and analyzed. The simulation results show that the positioning accuracy increases with the decrease of the intensity. The positioning accuracy of the improved algorithm using the interpolated fingerprint database is improved by 21.5% compared with that using the initial fingerprint database, and the computational complexity is greatly reduced.