Introduction: Image classification stands as a pivotal undertaking within the domain of computer vision technology. Primarily, this task entails the processes of image augmentation and segmentation, which are executed by various neural network architectures, including multi-layer neural networks, artificial neural networks, and perceptron networks. These image classifiers employ distinct hyperparameters for the prediction and identification of objects. Nevertheless, these neural networks exhibit susceptibility to issues such as overfitting and a lack of interpretability when confronted with low-quality images. Objective: These limitations can be mitigated through the adoption of Quantum Computing (QC) methodologies, which offer advantages such as rapid execution speed, inherent parallelism, and superior resource utilization. Method: This approach aims to ameliorate the challenges posed by conventional Machine Learning (ML) methods. Convolutional Neural Networks (CNNs) are instrumental in reducing the number of parameters while preserving the quality of dataset images. They also possess the capability to automatically discern salient features and maintain robustness in noisy environments. Consequently, a novel approach known as Deep Revamped Quantum CNN (DRQCNN) has been developed and implemented for the purpose of categorizing images contained within the Fashion MNIST dataset, with a particular emphasis on achieving heightened accuracy rates. Results: In order to assess its efficacy, this proposed method is systematically compared with the traditional Artificial Neural Network (ANN). DRQCNN leverages quantum circuits as convolutional filters with a weight adjustment mechanism for multi-dimensional vectors. Conclusions: This innovative approach is designed to enhance image classification accuracy and overall system effectiveness. The efficacy of the proposed system is evaluated through the analysis of key performance metrics, including F1-score, precision, accuracy, and recall
