The detection of Carrot Disease (CD) remains significantly impact by agricultural productivity and market quality. Therefore, early and accurate detection of diseases in carrot is essential for maintaining crop health and maximizing yield. This work proposes an automated Carrot Disease Detection (CDD) system that utilizes image processing and deep learning techniques. The Good and bad classification of Fresh CARROT dataset is fed to pre-processing stage where an Adaptive Wiener Filter (AWF) is used. AWF technique is to enhance image quality and reduce noise. This is followed by segmentation using a Kernel-Based K-Means algorithm to isolate diseased regions in carrot. The segmented carrot images are then fed to region-based feature extraction, capturing critical details relevant for disease identification. Finally, classification is performed using an Efficient Channel Attention Network (ECANet), to classify the carrot disease enabling high-accuracy predictions. Using Python software, the proposed framework achieved higher accuracy, F1-score, precision and recall of 94%, is accomplished when compared to other techniques.