Inspect Characteristics of Rice via Machine Learning Method

For more than half of humanity rice is life. Therefore, assessing the quality of rice in fast, accurate and objective methods has attracted a lot of attention of rice producers and processors. Unfortunately, the current inspection methods that focusing on the computer vision to inspect the characteristics of rice are either cost expensive (e.g., it needs extra sensors to assist photography) or need to be significantly improved in practice (e.g., framed object incorrectly). In this paper, we make an in-depth study of the characteristics of rice and explore an alternative direction to use machine learning methods to inspect them through photos taken by cell phones. To be exact, we develop a new mathematical variation formula and a new area calculation formula, combining clustering methods, to inspect the main characteristics of rice both statically and dynamically. The effectiveness of our approach is very visible no matter what the type of rice it is, which is shown by comprehensive experiments on four typical types of rice datasets. Moreover, We cooperate with one of the world’s largest home appliance manufacturers, applying the rice characteristics extraction approach to produce smart electric rice cookers, thus improving the quality of life for millions of people. Keywords— characteristics of rice; machine learning; digital image processing; IoT

[1]  Jens Michael Carstensen,et al.  Using Multispectral Imaging for Spoilage Detection of Pork Meat , 2013, Food and Bioprocess Technology.

[2]  Weiping Zhu,et al.  Spatial co-training for semi-supervised image classification , 2015, Pattern Recognit. Lett..

[3]  Weiping Zhu,et al.  Learning visual codebooks for image classification using spectral clustering , 2017, Soft Computing.

[4]  Jörg Weule,et al.  Non-Linear Gaussian Filters Performing Edge Preserving Diffusion , 1995, DAGM-Symposium.

[5]  Nitin Afzulpurkar,et al.  The Appropriate Algorithms of Image analysis for Rice Kernel Quality Evalution , 2006 .

[6]  Moshe B. Fuchs.  Topological structural analysis , 1997 .

[7]  Roberto Manduchi,et al.  Bilateral filtering for gray and color images , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[8]  Lihong Xie,et al.  Simultaneous Determination of Multi Rice Quality Parameters Using Image Analysis Method , 2014, Food Analytical Methods.

[9]  Ping Chen,et al.  On Improving the accuracy with Auto-Encoder on Conjunctivitis , 2019, Appl. Soft Comput..

[10]  T. F. Schatzki,et al.  APPLE WATERCORE SORTING SYSTEM USING X-RAY IMAGERY: I. ALGORITHM DEVELOPMENT , 2000 .

[11]  Saeid Minaei,et al.  Potential Applications of Computer Vision in Quality Inspection of Rice: A Review , 2015, Food Engineering Reviews.

[12]  P. Wallin,et al.  Foreign body prevention, detection and control , 1998 .

[13]  Lianyong Qi,et al.  Privacy-Aware Multidimensional Mobile Service Quality Prediction and Recommendation in Distributed Fog Environment , 2018, Wirel. Commun. Mob. Comput..

[14]  Yang Tao,et al.  DETECTING STEM AND SHAPE OF PEARS USING FOURIER TRANSFORMATION AND AN ARTIFICIAL NEURAL NETWORK , 2003 .

[15]  David C. Hoaglin,et al.  Some Implementations of the Boxplot , 1989 .