论文部分内容阅读
温室大棚在蔬菜培育中有着广泛应用,在高效生产的同时,除草问题亟待解决。该设计采用一种改进型的人工神经网络算法应对大棚作物苗期杂草识别,通过对遗传算法的神经元参数的优化,以减少错误的发生次数。结果表明:与采用径向基核函数的支持向量机算法相比较,改进型人工神经网络算法识别正确率更高,达到94%以上,可为进一步的除草机器人开发提供技术支持。
Greenhouse in greenhouse cultivation has a wide range of applications, while efficient production, weeding problems need to be solved. The design uses an improved artificial neural network algorithm to deal with greenhouse crop seedling weed identification, through the optimization of neuronal parameters of genetic algorithm to reduce the number of false occurrences. The results show that compared with the support vector machine (SVM) algorithm based on radial basis function, the improved artificial neural network algorithm has a higher correct recognition rate (94%), which can provide technical support for the further development of weeding robot.