The heading stage of maize is one of the most important periods during its growth and development and indicates the start of its pollination. In this regard, an automated method for maize tassel detection is highly important for precision agriculture. However, the recognition of maize heading stage mainly relies on artificial observation. This method presents some limitations, such as its high cost and subjectivity. This work proposes a novel method for automatic tassel detection. In the algorithm, a color attenuation prior model was used by modeling the scene depth of saturation graph to remove image saturation. The Itti visual attention detection algorithm was used to detect the area of interest. Finally, texture features were used to develop a classification model to eliminate false positives. Pictures were taken by a commercial camera in 2 years and used to verify the stability of the algorithm. Precision, recall, and F1 measure were calculated to quantitatively assess and rate the algorithms. Experimental results show that the proposed method outperforms other existing methods and shows recall, precision, and F1 measure values of 86.82%, 94.81%, and 90.64%, respectively. The results indicate that this proposed method can effectively detect maize tassels in field images and remain stable over time.