The work aims to explore children's perception preference of outdoor landscape color space in kindergartens, and propose a scientific and effective color design method for children's activity space. In this work, 25 kindergartens in Fuzhou were selected as the research objects. An color characteristic index system was established from the color composition and color space configuration. The p5. js software was used to develop programs to convert pictures into HSV space for segmentation and quantification. The ArcGIS10.2 and Fragstats4.2 software were used to calculate the color index, and the SPSS software was used to analyze the correlation of color characteristic indexes. The six color indicators with strong influence on children's perception preference were C1 color evenness index>C4 contrast index>C2 color quantity>C5 natural color proportion>SHDI plaque diversity index>FRAC_Mean fractal dimension of MN patch shape. C1 and C4 changed nonlinearly with preference level, C2, C5, SHDI, FRAC_MN was proportional to the preference level; Four common factors of color characteristics were extracted, which were color evenness and color quantity factor F1, color contrast and color saturation factor F2, color patch sub dimension factor F3, color patch diversity and natural color factor F4, with the weight of F1> F2> F3> F4. Three high-quality color space models are summarized, which respectively have the color characteristics of strong color contrast, rich natural colors, and diverse color composition shapes, providing theoretical reference for the improvement of color quality of kindergarten outdoor landscape and color design of other children's activity spaces.