<p>Plant diseases often affect various parts of plants, including fruits, young branches, leaves, and buds, leading to potential damage or wastage of the fruit. To prevent the spread of these diseases to other plants, early identification and appropriate measures are essential. To address this challenge, a novel Wild horse boosted Convolution Attention Bidirectional model for Plant diseases often affect various parts of plants, including fruits, young branches, leaves, and buds, leading to potential damage or wastage of the fruit. To prevent the spread of these diseases to other plants, early identification and appropriate measures are essential. To address this challenge, a novel Wild horse boosted Convolution Attention Bidirectional model for classifying healthy and diseased plant leaf images. The method integrates a Convolutional Neural Network with an attention mechanism and a Bidirectional Long Short-Term Memory network, further optimized using the Wild Horse Optimizer to enhance classification performance. The convolutional neural network component incorporates a regularization parameter to minimize overfitting and reduce the training parameter count, improving generalization. Attention layers emphasize salient features, which are then processed by the Bidirectional Long Short-Term memory to capture temporal and contextual dependencies in the feature space. The results validated the proposed Wild horse boosted Convolution Attention Bidirectional model in plant disease detection with the accuracy of 97.55%.</p>