Use of microscopy to assess bran removal patterns in milled rice

Predictive modeling of rice milling degree for three typical Chinese rice varieties using interpretative machine learning methods

Brown rice over-milling causes high economic and nutrient loss. The rice degree of milling (DOM) detection and prediction remain a challenge for moderate processing. In this study, a self-established grain image acquisition platform was built. Degree of bran layer remaining (DOR) datasets is established with image capturing and processing (grain color, texture, and shape features extraction). The mapping relationship between DOR and the DOM is in-depth analyzed. Rice grain DOR typical machine learning and deep learning prediction models are established. The results indicate that the optimized Catboost model can be established with cross-validation and grid search method, with the best accuracy improving from 84.28% to 91.24%, achieving precision 91.31%, recall 90.89%, and F1-score 91.07%. Shapley additive explanations analysis indicates that color, texture, and shape feature affect Catboost prediction accuracy, the feature importance: color > texture > shape. The YCbCr-Cb_ske and GLCM-Contrast features make the most significant contribution to rice milling quality prediction. The feature importance provides theoretical and practical guidance for grain DOM prediction model. PRACTICAL APPLICATION: Rice milling degree prediction and detection are valuable for rice milling process in practical application. In this paper, image processing and machine learning methods provide an automated, nondestructive, and cost-effective way to predict the quality of rice. The study may serve as a valuable reference for improving rice milling methods, retaining rice nutrition, and reducing broken rice yield.

Breakpoint Planning Method for Rice Multibreak Milling

Excessive milling of rice kernels will result in nutrient loss and grain waste. To avoid grain waste, multibreak milling systems have been widely used in large-scale commercial rice mills. However, there is still no reasonable breakpoint planning method to guide the multibreak milling process. To construct a reasonable multibreak milling system, in this research, taking rice milling, a typical heterogeneous cereal-kernel milling process, as an example, the multivariate analysis method was used to comprehensively analyze the characteristic changes of milled rice during the whole milling process. A breakpoint planning method was established, including planning the number of breakpoints, determining the degree of milling or milling time corresponding to each breakpoint, and estimating the actual breakpoint to which the milled rice belongs. The verification results showed the rationality and high accuracy of the planning method. The presented work will help operators to plan the multibreak milling system of rice efficiently and objectively so as to significantly improve the commercial value of milled rice.

Evaluation of Rice Degree of Milling Based on Bayesian Optimization and Multi-Scale Residual Model

Traditional machine learning-based methods for the detection of rice degree of milling (DOM) that are not comprehensive in feature extraction and have low recognition rates fail to meet the demand for fast, non-destructive, and accurate detection. This paper presents a digital image processing technology combined with deep learning to implement the classification of DOM of rice. An improved multi-scale information fusion model of the InceptionResNet-Bayesian optimization algorithm (IRBOA) was constructed based on the Inception-v3 structure and residual network (ResNet) model. It enables to automatically extract more comprehensive features of rice and determine the DOM of rice. Additionally, the important hyperparameters in the model were tuned by the BOA to optimize the recognition rate of rice DOM. The results show the hyperparameters optimized using the BOA are those that would not be chosen in manual tuning. The classification precision of the IRBOA model reached 99.22%, 94.92%, and 96.55% for well-milled, reasonably well-milled, and substandard rice, respectively, with an average accuracy of no less than 96.90%. This model improved 7.41% over the traditional machine learning model and at least 1.35% over the fashionable CNN model with strong generalization performance. This method effectively completes rapid, non-destructive, and accurate intelligent detection of rice DOM, which can supply a reliable and accurate technical mean for rice processing enterprises to guide the rice processing process.

Marker-Assisted Backcrossing (MABc) to Improve Eating Quality with Thin Seed Coat and Aleurone Layer of Non-Glutinous Japonica Variety in Rice

Brown rice is composed of rice bran, pericarp, seed coat, and aleurone layers, and the rice bran layer contains a large number of substances useful for the human body, such as dietary fiber, α-tocopherol, α-tocotrienol, and vitamins. However, more than 90% of these substances are removed when polished, and white rice has the disadvantage of losing food-related ingredients, such as umami-related amino acids, when compared to the unpolished group. In this study, we tried to develop new breeding lines with a thinner seed coat and aleurone layer to provide high eating quality with softer chewing characteristics and processability in rice grain. We detected an SNP for foreground selection for the backcross population by comparing genome sequences between Samgwang and Seolgaeng and developed high eating quality brown rice breeding lines by applying marker-assisted backcrossing (MAB_C) breeding programs to backcross populations between Samgwang and Seolgaeng using KASP markers. SNP markers for foreground selection were identified to improve eating and processability through SNP mapping of Samgwang and Seolgaeng with SSIIa as a target gene in this study. Line selection according to genotype of KASP markers was successful in BC₁F₁ and BC₂F₁ generations, with the recurrent parent genome recovery ratio ranging from 91.22% to 98.65%. In BC₂F₁ seeds of the selected lines, thickness of the aleurone layer was found to range from 13.82 to 21.67 μm, which is much thinner than the 30.91 μm of the wild type, suggesting that selection by MABc could be used as an additional breeding material for the development of highly processed rice varieties. These lines will be useful to develop new brown rice varieties with softer chewing characteristics and processability in rice grain.

Effects of cultivars and fertilization levels on the quality of rice milling: A diagnosis using near-infrared spectroscopy, X-ray diffraction, and scanning electron microscopy

Cultivars and fertilization levels influence rice productivity and can be associated with grain quality. Thus, it is possible to make decisions regarding the choice of cultivars and application of fertilizer levels based on the type of milling, a necessary post-harvest process that may minimize the nutrient load in the grains and result in loss in quality. This study relates the physicochemical composition and morphological quality of brown and polished milled rice grains, cultivar types, and different levels of soil fertilization using near-infrared spectroscopy analysis, X-ray diffraction and scanning electron microscopy. Statistical tools were used to test the various treatments and identify the relationship between factors and variables. A high fertilization level is related to increasing crude protein composition and starch for cultivar IRGA 431 CL associated with polished rice. However, the combination of cultivar IRGA 424 RI and brown rice demonstrated a higher grain resistance, and different percentages of whole, chalky, and damaged rice. The correlation between ash × crude protein and starch × crude fiber was found to be positive for brown rice and negative for the polished rice. Further, an increase in starch content was inversely proportional to the ash content, whereas an increase in crude protein was inversely proportional to the low-fat content in milled rice. The crystalline characteristics of rice starch were preserved at high fertilization levels associated with polished grains that demonstrated high starch content. Polished grains, however, showed more pores and cavities, and consequently greater permeabilities in the surface. It is recommended that batches of grains produced from cultivar IRGA 431 CL with high levels of fertilization be subjected to polished rice milling to achieve high protein and starch quality. However, grains from cultivar IRGA 424 RI with high levels of fertilization are recommended for brown rice milling owing to the high percentage of physical defects observed.