针对椭圆形农产品的分级问题,采用最近邻分类算法和随机Hough变换理论,对哈密瓜这类椭圆形农产品的大小分级方法进行研究。结果表明:1)哈密瓜边缘轮廓近似椭圆形,所测出的长短轴半径,可以作为椭圆形哈密瓜大小分级的新标准;2)通过试验测定,对于白色背景的哈密瓜,最近邻分类算法可以提取出较为完整的边缘轮廓;3)随机Hough变换可以在边缘轮廓不完整且有随机噪声干扰的情况下,检测出任意曲率的哈密瓜边缘轮廓的近似椭圆;4)与椭圆形哈密瓜的半径的真实值相比,本改进算法识别值的相对误差小于6%;5)当哈密瓜处在不同倾斜状态时,如0°、45°、90°、135°,本改进算法仍可以准确测得其长短轴半径。本改进算法还可以推广到其他椭圆形和类椭圆形农产品的大小分级中。 Aiming at the classification problem of oval agricultural products, the nearest neighbor classification algorithm and random Hough transform theory are used to study the size classification of oval agricultural products. The results showed as follows: 1) The contour of the cantaloupe was approximately elliptical, and the measured radius of the major and minor axes could be regarded as the new standard for the size classification of oval cantaloupe. 2) By the test, the nearest neighbor classification algorithm could be extracted for the cantaloupe with white background 3) random Hough transform can detect the approximate ellipse of cantaloupe edge contour with arbitrary curvature under the condition of incomplete edge contour and random noise interference; 4) with the true value of the radius of elliptic melon The relative error of the improved algorithm is less than 6%. 5) When the melon melons are in different tilted states, such as 0 °, 45 °, 90 °, 135 °, the improved algorithm can still accurately measure the short-axis radius . The improved algorithm can also be extended to the size of other oval and oval-shaped agricultural products.