<p>Color extraction accuracy and color matching efficiency are prerequisites for the application of computer vision computing in digital collectibles, smart design, and other fields. In response to this issue, this study proposes a color extraction and artistic matching model based on silhouette coefficient method and eye tracking technology. It improves its performance by optimizing color clustering and color matching evaluation strategies, and finally sets up experiments for verification. The results showed that in the simulation running experiment, the average color extraction accuracy of the research model before and after training was 0.36 and 0.95, respectively. Moreover, the average peak signal-to-noise ratio and structural similarity index of color extraction by the model were 20.86dB and 0.671, and the average efficiency of color matching was 97.35%. When the model lost the silhouette coefficient method and eye tracking technology, its peak signal-to-noise ratio, structural similarity index, and color matching efficiency decreased by 0.64dB, 0.020, and 7.36%. In addition, in actual model performance testing, the research model was able to accurately extract the main colors of the image without introducing impurities. In addition, its peak signal-to-noise ratio and mean absolute deviation for color were 19.66dB and 0.40dB, while its structural similarity index and mean absolute deviation were 0.658 and 0.004, respectively. Finally, the average score of the color matching scheme generated by the model was 9.00, which was 0.63 higher than the original image. In summary, the proposed color extraction and artistic matching model can improve the accuracy and quality of color extraction and matching, and achieve the expanded application of computer vision technology in image processing, digital devices, and other fields.</p>