Sugarcane Stalk Content Prediction in the Presence of a Solid Impurity Using an Artificial Intelligence Method Focused on Sugar Manufacturing

Detection of breakage and impurity ratios for raw sugarcane based on estimation model and MDSC-DeepLabv3

Broken cane and impurities such as top, leaf in harvested raw sugarcane significantly influence the yield of the sugar manufacturing process. It is crucial to determine the breakage and impurity ratios for assessing the quality and price of raw sugarcane in sugar refineries. However, the traditional manual sampling approach for detecting breakage and impurity ratios suffers from subjectivity, low efficiency, and result discrepancies. To address this problem, a novel approach combining an estimation model and semantic segmentation method for breakage and impurity ratios detection was developed. A machine vision-based image acquisition platform was designed, and custom image and mass datasets of cane, broken cane, top, and leaf were created. For cane, broken cane, top, and leaf, normal fitting of mean surface densities based on pixel information and measured mass was conducted. An estimation model for the mass of each class and the breakage and impurity ratios was established using the mean surface density and pixels. Furthermore, the MDSC-DeepLabv3+ model was developed to accurately and efficiently segment pixels of the four classes of objects. This model integrates improved MobileNetv2, atrous spatial pyramid pooling with deepwise separable convolution and strip pooling module, and coordinate attention mechanism to achieve high segmentation accuracy, deployability, and efficiency simultaneously. Experimental results based on the custom image and mass datasets showed that the estimation model achieved high accuracy for breakage and impurity ratios between estimated and measured value with R2 values of 0.976 and 0.968, respectively. MDSC-DeepLabv3+ outperformed the compared models with mPA and mIoU of 97.55% and 94.84%, respectively. Compared to the baseline DeepLabv3+, MDSC-DeepLabv3+ demonstrated significant improvements in mPA and mIoU and reduced Params, FLOPs, and inference time, making it suitable for deployment on edge devices and real-time inference. The average relative errors of breakage and impurity ratios between estimated and measured values were 11.3% and 6.5%, respectively. Overall, this novel approach enables high-precision, efficient, and intelligent detection of breakage and impurity ratios for raw sugarcane.

Color approach to the analysis of white crystal cane sugar for the detection of solid impurities

BACKGROUND

Sugar is consumed worldwide and so the quality control of sugar cane is necessary. Solid impurities are an inherent part of industrial sugar processing. Dark particles and adulteration with sand must be controlled. Sixty-four samples of white crystal cane sugar analysis in the presence of both kinds of impurities (dark particles and sand) were assessed using an affordable digital image system and a multivariate calibration strategy.

RESULTS

The quality parameters for the multivariate calibration models obtained to estimate sugar content were remarkable. Color descriptors from digital images allowed identification of different levels of sugar content for the following three ranges: 0-49.99 wt%, 50.03-78.99 wt%, and 82.99-100 wt%. The multivariate model using red (R), green (G), Blue (B), and luminosity (L) color descriptors showed low standard errors of cross-validation (SECV) and validation (SEV) of 7.63 and 6.01 wt%, respectively.

CONCLUSIONS

The method is affordable and reliable, and might aid quick screening in situations where access to a laboratory or instrumentation is restricted. © 2021 Society of Chemical Industry.

Chemometric approach in environmental pollution analysis: A critical review

With the ever-increasing global population and industrialization, it has become a call of the hour to start taking care of the environment to balance the ecosystem. For this, effective monitoring and assessment are required, which involves collecting and measuring environmental details, temporal and spatial readings of environmental data, and parameters. However, assessment of the environment is very tedious as it includes monitoring target analytes, identifying their sources, and reporting, which invariably implies that detailed environmental monitoring would be an intricate and expensive process. The traditional protocols in environmental measures are often manual and time demanding, which makes it further difficult. Moreover, several changes also occur within the environment, which could be chemical, physical, or biological, and since these environmental impacts are often cumulative, it becomes difficult to measure an isolated system. Furthermore, the chances of skipping significant results and trends become high. Also, experimental data obtained from the environmental analysis are usually non-linear and multi-variant due to different associations among various contributing variables. Therefore, it is implied that accurate measurements and environment monitoring are not using traditional analytical protocols. Thus, the need for a chemometric approach in environmental pollution analysis becomes paramount due to the inherent limitations associated with the conventional approach of analyzing environmental datasets. Chemometrics has appeared as a potential technique, which enhances the particulars of the chemical datasets by using statistical and mathematical analysis methods to analyze chemical data beyond univariate analysis. Utilizing chemometrics to study the environmental data is a revolutionary idea as it helps identify the relationship between sources of contaminations, environmental drivers, and their impact on the environment. Hence, this review critically explores the concept of chemometrics and its application in environmental pollution analysis by briefly highlighting the idea of chemometrics, its types, applications, advantages, and limitations in the environmental domain. An attempt is also made to present future trends in applications of chemometrics in environmental pollution analysis.

Artificial neural network employment for element determination in Mugil cephalus by ICP OES in Pontal Bay, Brazil

Fish are important sources of protein, making them very significant in the human diet. Although the consumption of this food is beneficial for health, it is essential that the product does not contain inorganic components above the limits recommended by the current legislation. Therefore, a method for determination of elements in fish (Mugil cephalus) samples was optimized. A simplex centroid mixture design with restriction was applied for optimization of the acid digestion of samples in an open system under reflux in order to evaluate the best ratio between the reagents HNO₃, H₂O₂ and H₂O. The results indicated that more intense analyte signals were obtained when a mixture containing 3.6 mL of HNO₃ (65% v/v), 0.4 mL of H₂O₂ (30% v/v) and 6.0 mL of H₂O was used. The accuracy of the method was assessed with a CRM of oyster tissue (NIST 1566b). The method presented relative standard deviations (RSDs) of 3.54%; 3.82%; 4.81% and 3.50% for Zn, Fe, Cu and S, respectively. The detection limits were 0.002 mg kg^-1 for Cu and Zn and 0.02 mg kg^-1 for Fe and S. The proposed method was applied for the determination of Zn, Fe, Cu and S in fish samples. A Kohonen Self-Organizing Map (KSOM) with K-means implementation was applied to better delimit the boundary between groups and the spatial and temporal influence on how concentrations of the chemical elements were perceived. To verify the separation, the Davies-Bouldin and Silhouette indices were used, obtaining 0.5374 and 0.8541, respectively, indicating satisfactory separation.