Document Type : Research Paper
Authors
Department of Agricultural Machinery Engineering,, Faculty of Agriculture, University College of Agriculture and Natural Resources, University of Tehran, Karaj, Iran
Abstract
Keywords
Main Subjects
EXTENDED ABSTRACT
Today, fried foods have gained significant popularity among people. Given the high consumption of fried foods, the health implications of using these products are important for many consumers. On the other hand, the healthiness of fried foods fundamentally depends on the quality of the frying oil and the method of frying the food. Fried products are delicious when cooked in oil, but they can also pose major health issues for consumers, including concerns related to calories, high cholesterol, cardiovascular diseases, changes in the structure and stability of the oil used, as well as an increase in trans fatty acids in repeatedly used oils. Exposure of oil to high temperatures and environmental air increases oxidation and produces toxic substances. Additionally, the shelf life of fried food decreases due to the oxidation of the oil (Michalak-Majewska et al., 2018). The global trend towards healthy eating has led to the development of new technologies that can preserve or enhance the quality characteristics of fried foods (Abd Rahman et al., 2016; anonymous, 2023).With the emergence of certain changes resulting from frying products using various methods, including deep frying and air frying, examining these frying methods can provide useful information for consumers regarding the appropriate choice of frying technique. This can significantly assist in satisfying consumers of fried food products. Analyzing the differences between various frying methods has been challenging for consumers. On one hand, deep frying offers better flavor and aroma, while on the other hand, air frying results in less oil absorption and improved quality of the fried product. These advantages and disadvantages of the two frying methods underscore the necessity for a scientific comparison between them. Many activities have been developed using machine vision technology in agricultural sectors (Kamilaris et al., 2018). This is because machine vision systems not only recognize the size, shape, color, and texture of objects but also provide numerical features of the objects or scenes being imaged. Developing a machine vision-based system to examine deep frying and air frying of mushrooms allows for a more precise evaluation of the visual characteristics and final quality of the food. This automated and comprehensive approach enables deeper analysis compared to traditional assessment methods. Using machine vision systems, qualitative and quantitative data can be simultaneously collected and analyzed. This feature allows for a more detailed examination of the two frying methods for mushrooms.The goal of this research is to utilize machine vision to study the frying of mushrooms using both deep frying and air frying methods. The innovation of this study lies in transforming the qualitative properties of frying into measurable quantitative values. By using image processing to identify brown and yellow color regions and calculating statistical parameters for these areas, the percentage of frying of the mushrooms can be quantified. This approach enables consumers to express the quality of frying in a quantitative and comparable manner, leading to better choices regarding frying methods.Ultimately, this research addresses the challenge of making a better choice between deep frying and air frying by presenting results from frying mushrooms using both methods. While some studies have focused on the nutritional impacts of frying methods, this research delves deeper into the relationship between visual quality and nutritional value.
In this study, 3 kg of fresh button mushrooms were purchased from Delsa Alborz Company. To continue the experiment, the mushrooms were weighed and grouped according to frequency. The number and weight of each in each group were recorded, and for each experiment, 3 mushrooms were considered according to the grouping based on frequency. The frequency of the number of mushrooms was approximately 9, 13, and 17 grams. According to this weighing, these three approximate weights were considered the grouping criteria, with each group having a weight difference of approximately 4 grams compared to the previous group. In order to show the apparent changes based on the different weights after frying the mushrooms, one mushroom was selected from each group in each experiment. This reduced the total weight difference. (The number of tested mushrooms was 12 for each temperature, and a total of 72 for both frying methods, depending on the repetition.) After grouping, each of the grouped groups was gently washed. After washing, the mushrooms were dried and prepared for frying. 4 replicates were considered for analysis and better results at each temperature and time level.
In this experiment, a hot air fryer was used for frying the mushrooms using the air frying method. The Tefal hot air fryer (model EY 801, manufactured in China with a power consumption of 1650 watts) was selected for hot air frying. The process of frying the mushrooms using the air frying method was carried out at temperatures of 170, 180, and 190 degrees Celsius, with three time levels of 10, 8, and 6 minutes (Basuny & Oatibi, 2016). These experiments were conducted with four repetitions for each temperature and time setting.
For frying the mushrooms, temperatures of 170, 180, and 190 degrees Celsius were used, with three time levels of 3, 2, and 1 minutes (Li et al., 2019). Four repetitions were considered for each temperature and time setting. To fry the mushrooms, the container was filled with 2 liters of oil. Once the oil reached the desired experimental temperature, the mushrooms were placed into the oil-filled container of the device. After frying at each specified temperature and time level, the weight of each mushroom was recorded.
The imaging process begins. For imaging, uniform lighting and photography settings are applied, and the sample is positioned for capturing images. The sample placement is rotating, allowing for imaging from various angles. The capability of capturing a series of images is utilized by holding down the volume up button on the mobile device for better imaging of the mushrooms. The mushrooms are placed on a rotating platform with the side that has the most details facing the camera, and imaging is performed from the top area to ensure that the maximum details are obtained.
After preprocessing, PCA is used to analyze the resulting data. PCA is a dimensionality reduction technique that help project data into a lower-dimensional space while preserving the maximum variance. This method can aid in reducing computational time and improving the performance of the proposed image processing models.
LDA works to find a linear combination of features that maximizes the separation between two classes (frying methods). This algorithm calculates the mean of the features for each class and performs the separation based on between-class and within-class variances.
After data reduction, the SVM method can be employed. The SVM model learns the features between images of both frying methods. Once the model is trained, it can be used to predict the classification of new images.
This system calculates the percentage of frying numerically by identifying brown and yellow regions in images, computing statistical parameters, and weighting them through calculations in a mathematical model. This feature allows for a more precise comparison of the frying quality across different methods. Overall, this research demonstrates the high effectiveness of using machine vision systems to detect differences between deep frying and air frying methods. It can serve as an effective tool for developing intelligent systems that assess the degree of frying and enhance the quality evaluation processes of food products. These results can assist both producers and consumers in selecting the best frying methods and improving the final quality of fried foods, ultimately leading to enhanced health and satisfaction for consumers.
Existing research has mostly been theoretical or conducted on a small scale, with less focus on industrial and commercial applications. This study and the expansion of research in this field can lead to the development of a practical system to create a standardized tool for the food industry, helping producers improve the quality of their fried products. Additionally, the implementation of this system and the expansion of research in this area could establish new standards for evaluating the quality of fried foods, ultimately enhancing the quality and safety of food products.
The use of modern methods such as machine vision systems holds particular appeal for consumers. Consumers are always looking for fried products that have the right texture, acceptable appearance, and are also satisfactory in taste and aroma. In air frying mushrooms, the mushrooms exhibit a more cohesive texture due to less moisture loss compared to those fried using the deep frying method. In deep frying, the texture of the mushrooms showed more wrinkling, and in terms of color, deep frying caused the mushrooms to darken towards a deeper yellow and brown color. In contrast, mushrooms fried using the air frying method had a yellowish hue, and in terms of appearance, the oil absorption in deep-fried mushrooms was high, to the extent that this amount of oil can be detrimental to the health of consumers and significantly affects the flavor of the mushrooms.
One of the key innovations of this research is the ability to quantitatively assess the doneness of mushrooms through color and texture analyses. This system calculates the percentage of frying numerically by identifying brown and yellow regions in images, computing statistical parameters, and weighting them through calculations in a mathematical model. This feature allows for a more precise comparison of the frying quality across different methods.
Donya Farajzadeh: Writing – original draft, Methodology, Data curation, Software, Formal analysis.
Seyed Saeid Mohtasebi: Review and editing, conceptualize, supervise, and manage project administration.
Mahmoud Soltani Firuz: review and editing, Formal analysis, Investigation, Validation.
Erfan Dehghan Banadaki: Writing, review and editing, Formal analysis, Data curation, Validation.
Data available on request from the authors. All the data used in this original research are presented throughout the text and in the form of Tables and Figures.
Thank you to all those who have collaborated in this research
The author declares no conflict of interest.
)