Due to the diversified development trend of modern architectural design, the functional design and application of houses are being presented in the interior space color layout of houses with its many advantages of interactivity, comprehensiveness, multi-functionality, and personalization. With this as the starting point, this paper analyzes the psychological perception of survey respondents on the same spatial color through big data, converts the perception into EEG signal for decoding, and after the decoding process of EEG data pre-processing, feature extraction, feature identification, and classification, calculates the significance of each element in the time-frequency matrix, which can get a homo-dimensional mask matrix. Then the conditional random field model is established on the random field theory to get the parameters of the model. Finally, the parameters of the model are obtained by maximizing the following entropy function, which is brought into the Lagrangian operator to obtain the pairwise Lagrangian operator. Finally, the EEG signal is decoded to realize the self-control training of color perception under different mentalities. The experimental results showed that by performing the intervention test on self-control and color insight, the mean value of the total self-control score in training was 61.99±11.45, and the intervention effect had stability. Therefore, improving self-control ability and forming correct tendency psychological perception plays a vital role in the color design of architectural space.