基于Android手机平台的玉米叶片含氮量无损检测

发布时间：2018-03-26 14:28

  本文选题：玉米叶片　切入点：含氮量　出处：《农业机械学报》2017年09期

【摘要】：为了提供一种玉米叶片含氮量无损快速检测方法,分析了玉米叶片的颜色特征参数与含氮量的关系,并基于Android手机平台开发了玉米叶片含氮量检测软件。首先获取包含被测玉米叶片与标定色块组的图像,利用标定色块对图像色彩进行校正,以减小外界光照等因素对图像色彩造成的失真。进而进行图像分割、图像平滑和颜色特征信息提取等处理,分析了各颜色特征参数与玉米叶片含氮量的关系,发现绿光标准化值与含氮量之间线性关系最好。应用Java语言和OpenCV计算机视觉库在Android手机平台上实现了玉米叶片的图像获取、图像处理和查看结果等功能。实验结果表明,该方法对玉米叶片含氮量的绝对测量误差为-0.40%~0.35%,均方根误差为0.20%,从采集图像到给出结果所用时间小于10 s。
[Abstract]:In order to provide a fast and nondestructive method for the detection of nitrogen content in maize leaves, the relationship between the color characteristic parameters and nitrogen content in maize leaves was analyzed. Based on the Android mobile phone platform, the detection software of nitrogen content in maize leaves is developed. Firstly, the images of maize leaves and calibration blocks are obtained, and the color of the images is corrected by the calibration color blocks. In order to reduce the distortion of image color caused by external illumination and other factors, image segmentation, image smoothing and color feature information extraction were processed, and the relationship between each color characteristic parameter and nitrogen content in maize leaf was analyzed. It is found that the linear relationship between the standard value of green light and the content of nitrogen is the best. The functions of image acquisition, image processing and viewing results of maize leaves are realized by using Java language and OpenCV computer vision library on the platform of Android mobile phone. The experimental results show that, The absolute measurement error of nitrogen content in maize leaves is -0.40 and 0.35, and the root mean square error is 0.20. The time from collecting images to giving the results is less than 10 s.
【作者单位】： 西北农林科技大学机械与电子工程学院;
【基金】：公益性行业(农业)科研专项(201503137)
【分类号】：S126;S513

【相似文献】

中国期刊全文数据库 前1条

1 李国栋 ,高秀月;尿素的妙用[J];农家参谋;1996年02期

，

本文编号：1668241