Fuzzy Edge-Detection as a Preprocessing Layer in Deep Neural Networks for Guitar Classification

Abstract

Deep neural networks have demonstrated the capability of solving classification problems using hierarchical models, and fuzzy image preprocessing has proven to be efficient in handling uncertainty found in images. This paper presents the combination of fuzzy image edge-detection and the usage of a convolutional neural network for a computer vision system to classify guitar types according to their body model. The focus of this investigation is to compare the effects of performing image-preprocessing techniques on raw data (non-normalized images) with different fuzzy edge-detection methods, specifically fuzzy Sobel, fuzzy Prewitt, and fuzzy morphological gradient, before feeding the images into a convolutional neural network to perform a classification task. We propose and compare two convolutional neural network architectures to solve the task. Fuzzy edge-detection techniques are compared against their classical counterparts (Sobel, Prewitt, and morphological gradient edge-detection) and with grayscale and color images in the RGB color space. The fuzzy preprocessing methodologies highlight the most essential features of each image, achieving favorable results when compared to the classical preprocessing methodologies and against a pre-trained model with both proposed models, as well as achieving a reduction in training times of more than 20% compared to RGB images.

Enhancing Secret Data Detection Using Convolutional Neural Networks With Fuzzy Edge Detection

“…The conventional Sobel and Prewitt operators operate similarly. Each utilizes a 3x3 gradient operator within a local neighbourhood, as adapted from [36]. However, their distinction lies in the convolutional process, where they employ different masks.…”

“…Recently, several works have tried to improve the performance of steganalysis; nevertheless, few combine fuzzy logic and CNN. Referring to Subsection A of Section II, among the three main methods to identify the fuzzy edges in a digital image, namely Prewitt, Sobel, and morphological gradients, we chose to use the Prewitt method in this article because it showed a superior performance as referred to [36]. In [11] an algorithm has been proposed to detect the location of the steganographic data in digital images based on fuzzy correlation maps for classification based on the results of this work to detect steganography and also departing from the significance of the results that fuzzy logic has yielded in FIGURE 3.…”

“…The general approach's depiction, the "Input" part consists of input preprocessing which contributes to overcoming the issues of the training dataset quality and quantity, and the remaining parts contribute in classification performance by mainly addressing the issues of both the poor features' learnability and the low payload detection. digital images classification [21], [36], [58], this work has been illuminated to use fuzzy logic as a preprocessing operation for images to be fed in the CNN for binary classification of images in cover or steganography holding. The position of this work, unlike [11], is to use Prewitt fuzzy edges identification to preprocess the inquiry images that are then used for training, validating, and testing our CNN.…”

“…This alteration only involves the mask, as identified earlier in Subsection A of Section II and denoted through the relations in ( 4) - (10). Nevertheless, this study adopted the Prewitt method due to its superior performance, demonstrated in [36]. Our inference system comprises fuzzy type-I with Mamdani fuzzy inference (FIS), featuring a pair of input and a single output membership function with three fuzzy rules.…”

Multiclass Mask Classification with a New Convolutional Neural Model and Its Real-Time Implementation

“…Our research group has worked with convolutional neural networks with diverse goals, as we can find in [ 10 ], where a new CNN model in combination with image preprocessing and optimization algorithms was proposed for diabetic retinopathy classification. Additionally, in [ 11 ] they use a deep neural network model for guitar classification, including fuzzy edge detection to improve the accuracy. In [ 12 ] a new hybrid approach was proposed, using fuzzy logic integration in combination with a modular artificial neural network.…”