基于高光谱图像技术预测苹果大小

以新疆红富士苹果为研究对象,探讨应用高光谱图像技术和最小外接矩形法预测其大小的研究方法。提取苹果高光谱图像中可见红色区域受色度影响较小的713nm以及近红外区域793和852nm的3个波长图像,做双波段比运算处理。比较所得双波段比图像可知,852/713双波段比图像中背景和前景灰度对比度最大。对该图像做阈值分割以及形态闭运算去除果梗区域,使用8邻接边界跟踪法得到二值图像的轮廓坐标序列,采用最小外接矩形法求苹果的大小,与实测值建立回归方程。结果表明,基于高光谱图像技术采用波段比算法,结合最小外接矩形法,能够有效地检测苹果大小,预测值与实际值最大绝对误差为3.06mm,均方根误差为1.21mm。

The Size Detection of Apple Using Hyperspectral Imaging Technology

The study investigated the size prediction performance of Xinjiang Fuji apple using hyperspectral imaging technology and Minimum Enclosing Rectangle method.Firstly,the three band images at 713nm in the visual red region and at 793 and 852nm in the NIR region,which were less affected by the color,were extracted from the apple hypersepctral images.And the operation of double band ratio were used to obtain band ratios images.Comparing with the gray diffe rence between the background and foreground of band ratios images,the gray contrast of 852/713 ratio mage is largest.After segmenting 852/713 ratio using threshold and removing stem’s region of binary image by morphological closing operation,the 8-Connected Boundary tracking method was used to get the edge sequences.The Minimum Enclosing Rectangle method was subsequently used to predict the size of the apple.Finally the correlation was established using predicted and actual value of the apple size.The results display that the hyperspectral imaging technology with the band ratio operation and the minimum enclosing rectangle method can be used to predict the size of Red Fuji apple,the maximum absolute error between the predicted and actual value of apple size was 3.06mm,and root mean square error was 1.21mm.