Crop and fruit production are essential for human being. But certain plant diseases destroy the plants and effects production. Plants have also the ability to keep a healthy environment; plants play a critical role in protecting the ecology and environment of the planet. The leaves are essential for distinguishing a plant because of their close closeness throughout the entire year. Between planting and harvesting, plant diseases may disrupt the leaf, causing a significant loss in crop output and financial gain. Determining leaf disease is therefore a challenging job in agriculture. However, it calls for a substantial staff, more processing time, and thorough familiarity with plant diseases. For both leaf type classification and disease detection, an optimization based deep learning (DL) method is used to solve this issue. Anisotropic filtering is used as a preprocessing step before Mask-R-CNN is used to segment the leaf picture. Here, disease detection and classification of leaf variety are multiclassified. Squeeze Net with Adaptive Snake Optimization (AdSnO) is utilized to classify leaf types. The deep QNet with the devised AdSnO is used in the detection of disease procedure. AdSnO-SqueezeNet strategy shown better outcomes in classification procedure with accuracy 0.92.