Plant disease identification acts a key role in maintaining agricultural productivity and ensuring food security worldwide. The continuous advancements in machine learning and computer vision technologies have directed to a remarkable transformation in the methods and tools engaged for this purpose. This survey article offers an in-depth analysis of the current and potential future directions in detecting plant diseases. The review involves a range of techniques and technologies, encompassing both traditional and modern methods containing machine learning, deep learning and deep transfer learning. The survey focus on plant diseases identification through the use of image processing, feature extraction, classification algorithms, and transfer learning. Additionally, challenges and opportunities within this domain are considered. Alongside proposed future research directions aimed increasing the accuracy, efficiency, and expandability of plant disease recognition systems. Furthermore, the document serves as a beneficial resource for individuals involved in the advancement of plant disease control and agricultural sustainability.