基于LoG算法的水果热成像检测

韩亚辉; 王琢; 刘佳鑫

基于LoG算法的水果热成像检测

韩亚辉^1,,
王琢^{1, 2, ,},
刘佳鑫^{1, 2}

1.
东北林业大学机电工程学院, 黑龙江哈尔滨 150040
2.
东北林业大学林业人工智能研究院, 黑龙江哈尔滨 150040

基金项目:

中央高校基本科研业务费专项资金资助 2572019CP21

详细信息

作者简介:
韩亚辉（1996-），硕士研究生。研究方向：图像处理。E-mail: 2962592004@qq.com

通讯作者:
王琢（1979-），硕士，副教授。研究方向：林业工程自动化、检测与测试技术。E-mail: wangzhuo@nefu.edu.cn

中图分类号: TN219
计量
- 文章访问数: 287
- HTML全文浏览量: 50
- PDF下载量: 39
- 被引次数: 0
出版历程
- 收稿日期: 2020-08-06
- 修回日期: 2020-12-04
- 刊出日期: 2021-07-01

Fruit Thermal Imaging Detection Based on Laplacian of Gaussian Algorithm

HAN Yahui^1
,,
WANG Zhuo^{1, 2
, ,},
LIU Jiaxin^{1, 2}

1.
College of Mechanical and Electrical Engineering of Northeast Forestry University, Harbin 150040, China
2.
Institute of Forestry Artificial Intelligence of Northeast Forestry University, Harbin 150040, China

摘要

摘要: 传统的水果分级与损伤检测大多采用感官评定的方法，随着计算机视觉技术的发展，计算机视觉检测分级技术发展迅速。研究中针对解决水果损伤部位检测的问题提出了一种利用图像处理技术对水果热成像损伤部位进行检测的技术方案。本方案采用Laplacian of Gaussian（LoG）算法对损伤部位进行检测，使用高斯卷积模板抑制噪声，通过设置不同的卷积核尺寸以及σ值获得不同的卷积滤波结果，加强了图像中损伤部位的色彩程度，进而更好地利用边缘检测技术获取损伤部位的边缘信息。采用具有局部损伤的苹果作为研究对象，选取有参考和无参考等5种评价方法，分析卷积过程对于损伤部位边缘检测的影响。结果表明，在水果热成像中LoG算法可以有效地检测水果的损伤部位，卷积核尺寸对于水果损伤部位边缘检测结果的影响远大于σ值，通过增大卷积核尺寸可以有效地加深损坏部分的边缘信息，研究为水果损伤区域检测提供了一种可行的解决方案。
- 水果损伤 /
- 热成像 /
- LoG算法 /
- 边缘检测
Abstract: Traditional fruit grading and damage detection mostly use sensory evaluation methods. With the development of computer vision technology, automatic computer vision detection and grading technology developed rapidly. To solve the problem of fruit damage detection, we propose a technical scheme for fruit thermal imaging damage detection using image processing technology. In this scheme, the Laplacian of Gaussian (LoG) algorithm was used to detect the damaged parts; a Gaussian convolution template is used to suppress noise. Different convolution filter results were obtained by varying the convolution kernel sizes and σ values to enhance the color degree of the damaged part in the image. Then, the edge detection technology was used to obtain the edge information of the damaged part. In the experiment, apples with local damage were selected as the research object, and five evaluation methods, including references and non-references, were selected to analyze the influence of the convolution process on the edge detection of damaged parts. The experimental results show that the LoG algorithm can effectively detect the damaged parts of fruits during thermal imaging, and the influence of the convolution kernel size on the edge detection results is far greater than the value of σ. By increasing the size of the convolution kernel, the edge information of the damaged parts can be effectively deepened. This study provides a feasible solution for fruit damage area detection.
- fruit damage /
- thermal imaging LoG algorithm /
- edge detection /

HTML全文

图 1 水果可见光图像

Figure 1. Visible light image of fruit

下载: 全尺寸图片幻灯片

图 2 损伤部位图像卷积结果

Figure 2. Convolution results of damaged image

下载: 全尺寸图片幻灯片

图 3 损伤部位边缘检测结果

Figure 3. Edge detection results of damaged parts

下载: 全尺寸图片幻灯片

图 4 MSE数据图

Figure 4. MSE data map

下载: 全尺寸图片幻灯片

图 5 PSNR数据图

Figure 5. PSNR data map

下载: 全尺寸图片幻灯片

图 6 SSIM数据图

Figure 6. SSIM data map

下载: 全尺寸图片幻灯片

图 7 Laplacian数据图

Figure 7. Laplacian data map

下载: 全尺寸图片幻灯片

图 8 方差数据图

Figure 8. Variance data map

下载: 全尺寸图片幻灯片

表 1 滤波结果量化值

Table 1. Quantized value of filtering results

MSE		σ＝5	σ＝10	σ＝20	σ＝30	σ＝40	σ＝50
	K=7	932.33	1033.55	1159.55	1157.57	1159.55	1159.21
	K=9	1850.98	2007.50	2525.38	2559.60	2561.29	2562.35
	K=11	3331.90	3340.24	4442.35	4816.12	4823.55	4826.10
	K=13	5057.46	5913.20	5876.27	5954.22	5985.96	6002.32
PSNR	K=7	18.4351	17.9875	17.4879	17.4953	17.4879	17.4892
	K=9	15.4568	15.1043	14.1075	14.0491	14.0462	14.0444
	K=11	12.9093	12.8930	11.6547	11.3038	11.2971	11.2948
	K=13	11.0915	10.4126	10.4398	10.3826	10.3595	10.3476
SSIM	K=7	0.961849	0.955389	0.948082	0.948056	0.948080	0.948086
	K=9	0.922108	0.926302	0.913651	0.911872	0.911722	0.911521
	K=11	0.834795	0.842830	0.793065	0.773563	0.773045	0.772895
	K=13	0.732744	0.712764	0.753693	0.755303	0.756641	0.756533
Laplacian	K=7	10.3903	11.1540	10.96140	10.3558	10.2464	10.2421
	K=9	9.63460	9.54343	10.63036	9.35341	9.34649	9.13412
	K=11	6.36219	6.24669	7.62676	7.15089	7.03636	7.00650
	K=13	6.66614	6.64652	7.12676	6.76413	6.57969	6.29320
Variance	K=7	1730.12	1736.66	1722.77	1666.19	1679.57	1679.31
	K=9	1426.23	1456.53	1367.94	1373.06	1366.66	1364.33
	K=11	1319.41	1397.31	1252.20	1274.32	1272.49	1272.55
	K=13	691.399	510.973	563.156	614.730	633.922	630.566

下载: 导出CSV

参考文献(17)

[1]	李广, 吴限鑫, 王建忠, 等. 水果功能性营养成分及其检测技术研究进展[J]. 农产品质量与安全, 2019(5): 75-80. doi: 10.3969/j.issn.1674-8255.2019.05.015 LI Guang, WU Xianxin, WANG Jianzhong, et al. Research progress on functional nutrients in fruit and related testing technology: a reviewv[J]. Quality and Safety of Agro-products, 2019(5): 75-80. doi: 10.3969/j.issn.1674-8255.2019.05.015
[2]	车远侠. 水果科普知识在学前教育中的作用——评《水果的秘密》[J]. 中国果树, 2019(6): 28. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGS201906051.htm CHE Yuaxia. The role of fruit science knowledge in preschool education—Comment on "The Secret of Fruits"[J]. China Fruits, 2019(6): 28. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGS201906051.htm
[3]	李承龙, 鲁明丽. 基于边缘检测技术的果品分级方法研究[J]. 常熟理工学院学报, 2018, 32(2): 78-82. doi: 10.3969/j.issn.1008-2794.2018.02.017 LI Chenglong, LU Mingli. Fruits Grading Algorithm Based on Detection Technology[J]. Journal of Changshu Institute of Technology, 2018, 32(2): 78-82. doi: 10.3969/j.issn.1008-2794.2018.02.017
[4]	李娟. 数字处理技术在果品分级中的应用——评《水果品质智能分级技术》[J]. 中国果树, 2019(6): 23. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGS201906046.htm LI Juan. Application of Digital Processing Technology in Fruit Grading——Comment on "Fruit Quality Intelligent Grading Technology"[J]. China Fruits, . 2019(6): 23. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGS201906046.htm
[5]	张妍. 果蔬农产品质量安全现状分析[J]. 农业科技与信息, 2019(22): 34-35, 39. doi: 10.3969/j.issn.1003-6997.2019.22.015 ZHANG Yan. Analysis on the Status Quo of Quality and Safety of Fruits, Vegetables and Agricultural Products[J]. Agricultural Science-Technology and Information, 2019(22): 34-35, 39. doi: 10.3969/j.issn.1003-6997.2019.22.015
[6]	肖付才, 靳雯雯. 数字图像边缘检测算法及在农产品检测中的运用[J]. 热带农业工程, 2019, 43(1): 133-135. https://www.cnki.com.cn/Article/CJFDTOTAL-RDZW201901046.htm XIAO Fucai, Jin Wenwen. Algorithm of Digital Image Detection and Its Application in Agricultural Product Processing[J]. Tropical Agricultural Engineering, 2019, 43(1): 133-135. https://www.cnki.com.cn/Article/CJFDTOTAL-RDZW201901046.htm
[7]	QIU S S, WANG J. Application of Sensory Evaluation, HS-SPME GC‐MS, E‐Nose, and E‐Tongue for Quality Detection in Citrus Fruits[J]. Journal of Food Science, 2015, 80(10): s2296-s2304. doi: 10.1111/1750-3841.13012
[8]	马康凌, 杨绍林, 刀静梅. 中国农产品检测技术现状与展望[J]. 云南农业科技, 2019(2): 60-63. doi: 10.3969/j.issn.1000-0488.2019.02.024 MA Kangling YANG Shaolin, DAO Jingmei. Current Status and Prospects of Agricultural Products Testing Technology in China[J]. Yunnan Agricultural Science and Technology, 2019(2): 60-63. doi: 10.3969/j.issn.1000-0488.2019.02.024
[9]	徐赛, 孙潇鹏, 张倩倩. 大型厚皮水果的无损检测技术研究[J]. 农产品质量与安全, 2019(5): 30-35, 48. doi: 10.3969/j.issn.1674-8255.2019.05.007 XU Sai, SUN Xiaopeng, ZHANG Qianqian. Research progress on nondestructive testing technology applied to large thick-skinned fruit[J]. Quality and Safety of Agro-products, 2019(5): 30-35, 48. doi: 10.3969/j.issn.1674-8255.2019.05.007
[10]	门洪, 陈鹏, 邹丽娜, 等. 基于主动热红外技术的苹果损伤检测[J]. 中国农机化学报, 2013, 34(6): 220-224, 229. doi: 10.3969/j.issn.2095-5553.2013.06.053 MEN Hong, CHEN Peng, ZOU Lina, et al. Apples'damage detection based on active thermal infrared technique[J]. Journal of Chinese Agricultural Mechanization, 2013, 34(6): 220-224, 229 doi: 10.3969/j.issn.2095-5553.2013.06.053
[11]	周建民, 周其显. 基于主动热成像技术的苹果早期机械损伤检测[J]. 农机化研究, 2010, 32(8): 162-165. doi: 10.3969/j.issn.1003-188X.2010.08.042 ZHOU Jianmin, ZHOU Qixian. Detection of Early Mechanical Danage inApple Based on Active Thermal Imaging[J]. Journal of Agricultural Mechanization Research, 2010, 32(8): 162-165. doi: 10.3969/j.issn.1003-188X.2010.08.042
[12]	马超. 基于多尺度多方向的图像边缘检测算法研究及其应用[D]. 开封: 河南大学, 2019. MA Chao. Research and Application of Image Edge Detection Algorithm Based on Multi-scale and Multi-direction[D]. Kaifeng: Henan University, 2019.
[13]	钟艺晶. 一种绝缘子串红外图像的特征提取方法研究[J]. 机电信息, 2020(21): 86-87. doi: 10.3969/j.issn.1671-0797.2020.21.042 ZHONG Yijing. Research on a Feature Extraction Method of Infrar-ed Image of Insulator String[J]. Mechanical and Electrical Information, 2020(21): 86-87. doi: 10.3969/j.issn.1671-0797.2020.21.042
[14]	黄玉蕾. 基于形态学滤波结合LOG算法的边缘检测[J]. 计算机测量与控制, 2019, 27(7): 257-260, 284. https://www.cnki.com.cn/Article/CJFDTOTAL-JZCK201907055.htm HUANG Yulei. Edge Detection Based on Morphological Filtering Combined with LOG Algorithm[J]. Computer Measurement & Control, 2019, 27(7): 257-260, 284. https://www.cnki.com.cn/Article/CJFDTOTAL-JZCK201907055.htm
[15]	张阳, 刘缠牢, 卢伟家, 等. 基于LoG算子的双滤波边缘检测算法[J]. 电子测量技术, 2019, 42(4): 95-98. https://www.cnki.com.cn/Article/CJFDTOTAL-DZCL201904017.htm ZHANG Yang, LIU Chanlao, LU Weijia, et al. Improved double fi-ltering LoG operator edge detection[J]. Electronic Measurement Technology, 2019, 42(4): 95-98. https://www.cnki.com.cn/Article/CJFDTOTAL-DZCL201904017.htm
[16]	贺萌. 基于自适应形态学的边缘检测及应用[D]. 长沙: 中南大学, 2013. HE Meng. Edge Detection Based on an Adaptive Morphology Algorithm and Application[D]. Changsha: Central South University, 2013.
[17]	SLANINA M, RICNY V. Estimating PSNR in High Definition H. 264/AVC Video Sequences Using Artificial Neural Networks[J]. Radio Engineering, 2008, 17(3): 103-108. http://www.oalib.com/paper/2490847