摘要
传统方法对水颜色识别的速度较慢、准确率不高。系统采用白色平行光束通过被测水和滤色片,分别照射在3块TCS230颜色传感器上,获得红、绿、蓝三基色(RGB)电压信号,并送入SCT89C516RD单片机进行数据处理。在算法上采用改进型BP神经网络算法,提高识别速度、增加识别级数和识别精度。该方法对海水、纯蓝墨水、红墨水的颜色进行实际识别。试验表明:该系统识别速度快,准确率高。
Traditional methods have many difficulties in identifing water color such as low speed and low accuracy.In this system,white parallel light passes through the water and filter and irradiates at 3 TCS230s color sensor respectively,so as to obtains red,green,blue trichromatic（RGB） voltage signal data for processing by microcontroller SCT89C516RD.This system adopted the improved BP neural network algorithm to improve the speed and accuracy and identify level.It carried out the experiments on identifing the colors of sea water,blue ink,red ink by this method and the results show that the system is high efficient and precise.
出处
《仪表技术与传感器》
CSCD
北大核心
2010年第7期102-104,共3页
Instrument Technique and Sensor
基金
广西自然科学基金重点项目(桂科自0832018Z)
广西科技攻关项目(桂科攻0895004-4)
广西科技厅攻关项目(0815003-11)
