Monitoring grocery inventory is important for effective supply chain management and waste reduction. Current grocery image classification models generally rely on RGB images but fail to capture complex variations in appearance and quality. This paper introduces MarketNet, a deep-learning model for simultaneous grocery image classification and segmentation using multi-input channels. It combines a multi-input convolutional neural network (CNN) and Vision Transformer (ViT) to improve classification and segmentation performance. The model incorporates multicolor input to learn richer and more discriminative features. The method was tested on a grocery image dataset, classifying items into predefined categories while simultaneously segmenting objects from backgrounds. Experiments show that MarketNet outperforms its baseline variants, achieving 98.87% Top-1 accuracy, 99.93% Top-5 accuracy, 94.54% mIoU, 96.73% accuracy, 96.81% precision, 97.58% recall, and 97.19% F1-score. This work supports improved inventory management, reduces product wastage, and enables real-time product tracking in grocery settings.