Non-invasive identification of apple sugar content based on convolutional neural networks

Yijia Zeng; Yuwei Cai; Hao Liu; Ruiquan Chen; Zeyu Xiao; Sihan Wu; Xiao Peng; Junle Qu

doi:10.1117/12.2654932

15 December 2022 Non-invasive identification of apple sugar content based on convolutional neural networks

Yijia Zeng, Yuwei Cai, Hao Liu, Ruiquan Chen, Zeyu Xiao, Sihan Wu, Xiao Peng, Junle Qu

Proceedings Volume 12478, Thirteenth International Conference on Information Optics and Photonics (CIOP 2022); 1247845 (2022) https://doi.org/10.1117/12.2654932
Event: Thirteenth International Conference on Information Optics and Photonics (CIOP 2022), 2022, Xi'an, China

Abstract

There is a high demand of novel monitoring methods for apple content measurement, especially sugar content (SC). Two traditional methods are widely used: one obtains sugar degree of apple juicy; the other identifies SC from intact apples by using near-infrared reflectance and optical fiber sensing techniques. The former is destructive and cumbersome. The latter requires expensive spectrometer equipment. Recently, deep learning has played an important role in image recognition. Convolutional Neural Network (CNN) has stronger capabilities of feature extraction and model formulation. Here, we have applied CNN into evaluate apple SC. Firstly, images of apples with SC in the range between 10 to 15 were sampled to generate data sets, which were used for data augmentation to generate larger data sets. In image processing, semantic segmentation was used to separate the target apple image from ambient noise. In the following training process, the extracted data sets were input into CNN-based deep learning model to provide the apple SC prediction model and the accuracy of prediction yield. After that, the network structure and hyperparameters were optimized to a satisfactory level, ensuring this apple sugar degree prediction model to achieve an accuracy of about 90% on the test set of apple images. Moreover, this CNN-based apple SC model was deployed on the mobile phone for achieving high portability. In conclusion, the CNN-based prediction method of apple sugar content has the advantages of non-invasive property, low cost, fast speed, high accuracy and flexibility, indicating great potential in practical applications of fruit industry.

Citation Download Citation

Yijia Zeng, Yuwei Cai, Hao Liu, Ruiquan Chen, Zeyu Xiao, Sihan Wu, Xiao Peng, and Junle Qu "Non-invasive identification of apple sugar content based on convolutional neural networks", Proc. SPIE 12478, Thirteenth International Conference on Information Optics and Photonics (CIOP 2022), 1247845 (15 December 2022); https://doi.org/10.1117/12.2654932

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
6 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Data modeling

Image segmentation

Convolution

Convolutional neural networks

Performance modeling

Feature extraction

Near infrared

Show All Keywords

Keywords/Phrases

Search In:

Publication Years