Due to its practical significance, the deep learning (DL)-based approaches are widely considered to solve a variety of the classification tasks based on the chosen digital images. Recently, the DL-based methods are widely adopted in variety of domains including the food item monitoring. This work aims to develop a DL-tool to monitor healthy/infected lettuce leaf (LL) from the digital images prepared with a digital camera. The LL is one of the common parts in most of the salads and sandwich and choosing the fresh LL is necessary. Hence, more care must be taken while selecting the LL during the food preparation. This research aims to develop a DL-based method to detect the healthy/infected LL based on the chosen leaf image. Different stages involved in this system includes; LL image collection and resizing, feature extraction using the MobileNet-variants, implementing 50% features reduction and serial features fusion, binary classification and threefold cross validation. In this work, the proposed DL-tool is developed using the MobileNet-variants and the performance of the proposed tool is confirmed using the SoftMax classifier with the individual and fused features. This study confirms that the proposed tool provides 98% accuracy when the fused-feature-based classification is executed.