At present, tobacco topping and sprout inhibition is mainly by artificial means, existing machinery can only be achieved ‘make it rigidly uniform’ . In order to improve work efficiency of tobacco topping and sprout inhibition, to reduce labor intensity，the machine vision technology is applied to tobacco topping and sprout inhibition operation, to realize intelligent detection of tobacco height. Firstly, the color characteristics of tobacco leaves and tobacco flowers within the visible light band were analyzed, and the dyeing cloth similar to green leaves was selected as image background by using Pantone which can reduce the extraction difficulty of tobacco flower. Then the common camera was used as the image acquisition device, and the industrial control computer was used for processing the color characteristics of the collected image. Based on single line adaptive threshold value judgment method and taking volume as a unit, R value and G value of different characteristics were used to find the location of tobacco flowers, according to the proportion relationship to determine the specific topping height. Finally, the industrial machine was combined with microcontroller to control the stepper motor and solenoid valve, so as to realize the real-time online tobacco topping and sprout inhibition. The test shows that the intelligent tobacco topping and sprout inhibition machine operation efficiency is 2~3 seconds per tree, tobacco flower detection accuracy is about 96％, topping accuracy rate is about 90%, sprout inhibition accuracy rate is about 80%, the machine can meet the design requirement, and can effectively improve the operating efficiency and operation effect.