Great advances in artificial intelligence and the Internet of Things have been made in recent years. This work develops an intelligent agricultural cultivation system to realize automated and intelligent farm cultivation management and operation. Experimental results demonstrate that the proposed system reduces water consumption by 50% and 30%, respectively, below those of fixed and threshold irrigation methods and predicts more accurately the amount of irrigation water required. The proposed dynamic detection transmission mechanism performs 97.8% fewer communications than are performed in timing transmission mode, effectively reducing overall energy consumption for transmission. Experimental results show that the proposed system reduces cultivation time by 40.94% below that required by the traditional cultivation method, and supports more effective management. Based on all of the experimental results, the proposed method supports convenient cultivation for growers.