H.266/Versatile Video Coding (VVC) is the latest international video coding standard to support high-definition video with resolutions from 4K to 8K and beyond. The compression ratio is higher than that of H.265/High Efficiency Video Coding (H.265/HEVC). In addition to the quad-tree partition structure in H.265/HEVC, the multi-type-tree (MTT) structure consisting of the binary tree and the ternary tree provides more diverse splits in H.266/VVC. It is also equipped with many new coding tools, which greatly increases the encoding time. This paper proposes a fast H.266/VVC intra coding algorithm based on visual perception analysis. According to the just-noticeable-difference in visual perception, the visually distinguishable pixels are extracted. Two intra coding tools are turned off conditionally. By using random forest classifiers of machine learning, the fast horizontal/vertical splitting decisions for the binary tree and the ternary tree are proposed. Under the All-Intra configuration, the experimental results show the proposed algorithm can save the encoding time by 47.51% with a BDBR of 1.454% on average. The proposed algorithm outperforms the previous research.