摘 要近年来，随着信息化进程的不断推进，图像处理技术已广泛应用于各个领域，成为信息时代一项不可或缺的技术。尤其近年来随着平板电脑和智能手机的普及，其高清拍照、二维码识别、在线视频、名片识别等功能深受大众欢迎，更加说明了图像处理技术的重要性。图像处理技术包含对图像的增强、分割、变换等，将这些方法用于水果图像的处理，就可以实现对水果的自动化检测分类。不仅提高了检测效率，还降低了劳动力。82801

本文比较系统的研究了水果的分类识别，其中包括对大小和缺陷的识别分类。在识别分类之前，先介绍了图像的预处理方法，包括图像的灰度化、图像平滑和图像增强。通过对不同处理方法进行对比，选择合适的方法处理图像。其次，水果的特征提取重点在于图像分割，因此选择合适的方法和阈值对图像进行分割显得尤为重要。在完成上述图像处理后，就是对水果大小的测量和缺陷的检测了。通过对三种大小测量方法的比较，选择误差较小的扫描法来获得果径。缺陷检测方面，利用不同区域灰度值之间的差别实现对缺陷区域的检测。由于缺陷检测是基于灰度图像处理的，相比于直接对彩色图像进行处理，该方法速度更快，效率更高。

毕业论文关键词：图像处理；水果分类；边缘检测；缺陷特征；

Abstract

In recent years, with the progress of information technology, image processing technology has been widely used in various fields, and become an indispensable technology in the information age。 Especially in recent years, with the popularity of smartphones and tablet computers, the high-definition camera, two-dimensional code recognition, online video, business card recognition function by popular and more shows the importance of image processing technology。 Image processing technology contains the image enhancement, segmentation, transform, and so on, these methods can be used for fruit image processing, you can achieve the automatic detection of fruit classification。 Not only to improve the detection efficiency, but also to reduce the labor force。

In this paper, a systematic study was made on the classification of fruits, which included the identification of size and defects。 Before the classification, the image preprocessing method is introduced, including image gray, image smoothing and image enhancement。 Through the comparison of different processing methods, select the appropriate method to deal with image。 Secondly, the key point of the feature extraction is the image segmentation, so it is very important to choose the appropriate method and threshold to segment the image。 After the completion of the above image processing, is the measurement of the size of the fruit and the detection of defects。 Through the comparison of the three measurement methods, the error of the scanning method is chosen to obtain the results。 Defect detection, using the difference between the gray value of different regions to detect the defect region。 Because the defect detection is based on the gray image processing, compared to the color image processing, the method is faster and more efficient。

Keywords: image processing; fruit category; edge detection; defect features

目 录

第一章 绪论 1

1。1 课题研究背景及意义 1

1。2 基于图像处理的水果识别国内外研究现状 2

1。2。1 国外研究现状 2

1。2。2 国内研究现状 3

1。3 本文结构及主要内容 3

第二章 图像处理基本知识