The work aims to propose a reverse color space conversion algorithm based on the cube inner contraction, in order to realize the conversion from CIELab color space to RGB color space. By selecting 4096 group modeling points and 512 group test points and taking the three-dimensional look-up table interpolation method as the theoretical basis, a method based on the cube inner contraction was designed to search the eligible feature points to replace the test points step by step, and then the RGB values of test points after reverse conversion were calculated. The conversion accuracy of the algorithm was evaluated by CIELAB 1976, CIE94 and CIE2000 chromatic aberration formula, and the chromatic aberration between the Lab value of the test point and the corresponding Lab value from the reversely converted RGB was calculated. The average chromatic aberrations of the test points were 2.07, 1.53 and 0.96 respectively. The distribution range of most chromatic aberrations was between 0 and 3, and the conversion accuracy of the algorithm was quite high. The result of test points after reverse conversion was ideal, but the accuracy in the green area and in the red low-brightness area could be further enhanced. The experimental results show that this algorithm can realize the conversion from CIELab color space to RGB color space accurately, quickly and effectively.