基于计算机视觉的水果分级检测系统的毕业设计.doc

基于计算机视觉的水果分级检测系统的设计 摘 要 计算机视觉应用于水果的品质检测，带来了许多方便。既可以提高检测的精度、准确度。又节省了大量的劳动力，让人们从繁重的人工检测工作中解脱出来。本文以苹果为研究对象，研究了计算机视觉技术应用于水果分级检测的基本理论和方法。 研究了苹果图像的预处理，包括平滑滤波、图像的灰度化以及图像的二值化。 研究了苹果的大小检测。先把苹果图像与背景分离，再计算出苹果图像的像素点数，通过预先测定出的一个像素点与真是面积的比值，进而算出苹果的真是面积，最后通过直径的大小来确定苹果大小等级。 研究了苹果的颜色检测，通过HIS颜色模型中的H分量来判定出苹果的着色面积，通过着色面积与苹果的大小做比，得出苹果的着色比，通过着色比来判定苹果颜色等级。 研究了苹果的缺陷检测。对苹果图像的灰度化，再通过用合适的阀值二值化图像确定出缺陷区域，在通过一些简单的运算得出缺陷的面积，通过缺陷的面积确定苹果的缺陷等级。 关键词：计算机视觉图像处理水果分级 he Design of Fruit Grading Detection System Based on Computer Vision ABSTRACT Computer vision applied to fruit quality inspection, brought a lot of convenience. Can enhance the detection accuracy. And save a lot of peoples labor from the heavy manual inspection work in earnest. So today I will introduce the basic theory and methods of a technology which can detect fruit ,this technology takes apple as the research object . Apple image preprocessing, including filtering, the grayscale of the image and the binarization of images. The size of the detection of apple. Departing apples image and background first, secondly ,calculate the number of the apple images pixels. Thirdly ,calculate the area of the apple in real through the predetermined ratio of a pixel area and its real area.Finally,determine the apples size class through diameter . The color of apple detected by its color model HIS.We use the component H in HIS model to determine apples colored area, through the ratio of the colored areas size and the apples size in real we can find out the color ratio. So we can determine the class of color through color ratio. The apple defect detection.We should make out the grayscale image of the apple at first, and then by using the appropriate threshold of the binary image to determine the defect area, so we can draw out the area of the detection through some simple operations .Finally we can determine the defect level of this apple through the area of detection. Key words: Computer vision, image processing, fruit grading 目 录 1绪论 1 1.1 研究的目的与意义