The pottery carved symbols excavated from Shuangdun site carry the cultural information of 7300 years ago, and their digital inheritance and innovative application are of great significance. This paper adopted deep learning and computer vision technology, constructs an image brightness feature extraction model based on the principle of perceptron, and intelligently recognized and extracted features of Shuangdun Carved Symbols through convolution operation, Harris corner detection, SIFT and SURF algorithms. The study established a complete algorithmic process including luminance value conversion, feature enhancement, thresholding and inversion operations, realizing the automated analysis and processing of symbol images. The results show that the overall user satisfaction of cultural and creative products reaches 4.0581 points, with the highest satisfaction of 4.1993 points in the innovativeness dimension. The market research found that 71.6485% of consumers pay more attention to the beautiful symbolism of the products, 76.6485% of purchasers use the cultural and creative products for souvenir collection, and 52.3485% of respondents prefer the national trend style design. Through the deep integration of intelligent algorithms and traditional cultural symbols, a complete technical system was established, covering symbol analysis to product design, which provides a feasible technical solution for the digital inheritance and innovative development of Neolithic cultural heritage.
