Tomato crops are vastly vulnerable to diseases, which reduce their yields and qualities adversely. Early diagnosis of these diseases is very important if the diseases are to be controlled adequately. The current research paper focuses on the identification of sophisticated techniques in tomato diseases diagnosis. To this end, we discuss and analyses various methods and algorithms that include the primitive approaches, leading to the machine learning approaches and recent deep learning models. This study identifies Image Processing and Pattern Recognition algorithm which has been found useful in early diagnosis of the diseases. Finally, we look at the use of sensors and IoT appliances for real-time measurements and improved data capture. The findings call for high quality datasets that would enhance the quality of images used in these models. Additionally, we present the drawbacks of current methods and suggest some ideas on how to improve the reliability and expand the capacities of tomato disease detection systems. Consequently, this systematic review shall be beneficial to intending and existing researchers and practitioners in the field, with a view of lending support to the formulation of resource-efficient and effective disease diagnosis frameworks for sustainable tomato production.