With the advancement of agricultural modernization, precision agriculture technology is increasingly valued. Weed identification and meristematic tissue localization are crucial for precise weed control, as they can effectively reduce the use of herbicides and improve weed control effectiveness. To meet the requirements for meristem center location in applications such as laser weed removal, a method for weed identification and meristem center location has been proposed. Firstly, weeds are identified using the designed convolutional neural network model. Based on weed identification, a series of matching templates are devised to locate skeleton intersections through matching techniques, ultimately pinpointing meristematic points. Experimental results indicate that the proposed method achieves an accuracy rate of 89.5% in weed identification. Following weed identification, the proposed method effectively determines the central point of the weed meristem, offering robust technical support for weed removal in precision agriculture.