Artificial neural network classifier for quality inspection of nuts

Abstract

In the food industry, automatic inspection of key ingredients as well as end products is gaining attention. Automatic detection of unhealthy ingredients on early stage of food production is becoming a vital task. In this paper, feature extraction of raw food ingredient's x-ray images and their classification is presented. X-ray images of pine and pistachio nuts are used as samples for this purpose. Statistical and texture features are extracted from each of original image samples as well as after applying edge detection on them. Texture features are calculated on global level and from co-occurrence matrices. Original features are used for classification independently, and in combination with edge features also. Artificial neural network is used as classifier. Texture features independently turned out better results with the classifier producing 0% and 6.8% false negative rate for pine and pistachio nuts respectively.

Power Outage Estimation: The Study of Revenue-Led Top Affected States of U.S

“…It has the advantage of capturing highly complexity in the data in the presence of outliers. The ANN is the other common classifier inspired by the biological brain, being used widely across all research areas [45][46][47][48]. It has the capability to adopt non-linear relationship between input and target.…”

“…The weights of the network are optimized after several forwardbackward passes by minimizing the difference between the actual output and the predicted output. The ANN has been extensively used for last two decades by the researchers across wide application areas [45][46][47][48].…”

Pattern Discovery from Big Data of Food Sampling Inspections Based on Extreme Learning Machine

“…A variety of methods and technologies have been studied, tested and/or implemented to analyze and utilize the data collected from food safety inspections, and many exciting results and conclusions have been obtained. Khosa and Pasero [6], [7] used an artificial neural network (ANN) as a classifier to predict at an early stage of processing or manufacturing whether important food ingredients, pine and pistachio nuts, are healthy. X-ray images of the nuts were used, and texture features were extracted from the images.…”

“…Logistic regression, like many other regression methods, is essentially linear regression; it is aided by some nonlinear transformations, and it can capture the nonlinear relationships between the dependent variable and causative (independent) variables. The ANNs in [1], [7] used a considerable number of nonlinear transformations to capture more detailed relationships. However, as demonstrated earlier, the learning speed of feed-forward ANNs is considerably slower than that of regression learning algorithms, which take the least squares method (LSM) as the core technique.…”

“…Fulfillments of these applications are helpful for boosting food quality level and improving food safety tracking. In order to implement these applications, however, a bunch of approaches are needed such as data preprocessing, statistical analysis, machine learning, and data mining [3,6,7].…”