Potato leaf diseases cause harm by reducing vegetative grade and yield, results to worldwide challenges. This is a primary cause of the scarcity of food and the rising expenses associated with food production. To detect the infections, an autonomous system built on machine learning evaluated against the most recent deep learning models, preferring to focus on more conventional methods. The present study introduced a Deep-learning based named Dual Channel Convolutional Neural Network (DC-CNN) model employs to detect the disease that are present in potato leaves. Plant Village Datasets are the input image's source data collection. To remove noise from input data, the data is pre-processed using the Kalman filter. Mask Region-based Convolutional Neural Networks, providing exact segmentation masks at the pixel level for each detected image, have been utilized in feature extraction. Furthermore, the images are classify using a Spiking Neural Network to determine whether the data given is normal or diseased. Finally, dual-channel convolutional segmentation is demonstrated, which enhances the performance of multimodal tasks and produces representation with attributes that are more valuable and reliable. The outcomes of the experiment prove the proposed DC-CNN strategy provides greater accuracy range of 99%, respectively.