Prediction of dry matter content of recently harvested 'Hass' avocado fruits using hyperspectral imaging

Hyperspectral technique combined with stacking and blending ensemble learning method for detection of cadmium content in oilseed rape leaves

BACKGROUND

Oilseed rape, as one of the most important oil crops, is an important source of vegetable oil and protein for mankind. As a non-essential element for plant growth, heavy metal cadmium (Cd) is easily absorbed by plants. Cd will inhibit the photosynthesis of plants, destroy the cell structure, slow the growth of plants, and affect their development and yield. It is necessary to develop a method based on visible near-infrared (NIR) hyperspectral imaging (HSI) technology to quickly and nondestructively determine the Cd content in rape leaves.

RESULTS

Two-layer estimation models were established by combining visible-NIR HSI with ensemble learning methods (stacking and blending). One layer used support vector regression, extreme learning machine, decision tree, and random forest (RF) as basic learners, and the other layer used support vector regression or RF as a meta learner. Different models were used to analyze the spectra of rape treated with five Cd concentrations to obtain the best prediction method. The results showed that the best model to predict Cd content was the stacking ensemble model with RF as the meta learner, with coefficient of determination for prediction of 0.9815 and root-mean-square error for prediction of 5.8969 mg kg^-1 . A pseudo-color image was developed using this stacking model to visualize the content and distribution of Cd.

CONCLUSION

The combination of visible-NIR HSI technology and the stacking ensemble learning method is a feasible method to detect the Cd content in rape leaves, which has the potential of being rapid and nondestructive. © 2022 Society of Chemical Industry.

MEDUSA: An open-source and webcam based multispectral imaging system

Multispectral imaging is at the forefront of contactless surface analysis. Standard multispectral imaging systems use sophisticated software, cameras and light filtering optics. This paper discloses the building of a customizable and cost-effective multispectral imaging and analysis system. It integrates a web camera, light emitting diodes (LEDs) lighting, a semisphere for even lightening, an open-source Arduino™ development board and a free Python application to automatically obtain and visually analyze multispectral images. The device is hereafter called MEDUSA and its optical performance was tested for repeated Imaging consistency, visible and near infrared band sensitivity and lighting evenness. Four proof of concept tests were run in order to understand the advantageous use of this system, as compared to a simple visual score of diverse samples. Each of three qualitative tests used sets of 12 LED band spectral images to analyze ink changes in a counterfeit bill, surface bruises on Hass avocado fruits and transient changes in petri dish grown bacterial colonies. A fourth test used single band imaging in a set of standard laboratory analyzed plant samples, to quantitatively relate a red band light reflectance to its nitrogen content. These tests indicate that MEDUSA made images may yield qualitative and quantitative spectral information unseen to the naked eye, suggesting potential use in currency counterfeit tests, food quality analyses, microbial phenotyping and agricultural plant chemistry. MEDUSA can be freely reproduced and customized from this research, making it a powerful and affordable analytical tool to analyze a wide range of subtle chemical properties in samples at industrial and science fields.

Prediction of the Physicochemical and Nutraceutical Characteristics of 'Hass' Avocado Seeds by Correlating the Physicochemical Avocado Fruit Properties According to Their Ripening State

Vegetal wastes are currently a source of pollution due to the excess of organic compounds in the environment. Seeds are the main by-product of the avocado industry and represent 16-22% of the total weight and it is considered a waste without applications. Despite the seed stands out for its high content of phenolic compounds, lack of knowledge regarding of the best processing state using non-invasive and short-time methods are required to take advantage of these nutraceutical compounds. This research aimed to find correlations from physicochemical analysis, color, hardness, and firmness of the whole avocado seeds with its nutraceutical properties as long as the ripeness increased, providing information for further industrial use of this waste. The results indicated that 'Hass' avocado fruit ripening positively correlates with the improvement of the physicochemical parameters involved in the fruit processing and the increase of nutraceutical compounds in the seed. The ripeness process decreased moisture (%) and hardness (N) parameters in the seeds (27.69 and 16.4%, respectively), facilitating its processing. Moreover, the ripening increased the antioxidant capacity by DPPH* around 7%, due to the concentration of phenolic compounds in the seed. Seed's phenolic compounds were positively correlated with the Hue angle at increasing ripeness, becoming a potential physicochemical indicator for the industry. The prediction of changes in nutraceutical compounds and physicochemical properties, as ripening occurred, may reduce analysis times, processes, and guidance to use avocado seeds as a by-product. These results facilitate the seed processing and open up opportunities for its use in the industry.