I. Introduction
Diseases have a significant impact on agricultural output across the world. The global agriculture economy suffers as a result of falling productivity. The apple tree is one of the world's most widely planted fruit plants-popular fruits throughout the globe. According to estimates, global production in 2018 was a banner year for apples in terms of production and consumption. However, the national average apple yield needs to catch up to the fruit's true potential. Apple's output could be higher due to ecological issues and inadequate post-harvest practices. Technology, decrease funding for fundamental research, a shortage of high-quality planting farming communities, socioeconomic limitations, etc. As much as they consume despite, their therapeutic value, apple trees are susceptible to several pests and illnessestiny creatures like bacteria. Apple is susceptible to several different illnesses. The anthrax-causing bacterium, a fungus called Gymnosporangium juniper Vir-powdery mildew (Powdered Fusarium oxysporum), Ginnane(Fireblight), scab (Venturia inadequacies), and scab The White-spotted Pygmy Seahorse (Podosphaera leucotricha). As a result, using fertilizers for trees properly is crucial. To aid farmers and avoid losses, early detection of such problems leaves more losses by doing the right things. When only conventional methods of diagnosis are used, sometimes farmers miss the window of opportunity to prevent plant diseases when they are first detected. Currently, Expert oversight is required, and no automated processes exist for such rapid identification. Regular use is a must. Inefficiency caused by a lack of automation costs reduces the taste of fruits and vegetables. The development of technology has led to a shift that is particularly popular in this field, including machine learning and other forms of soft computing. Canny Edge Detection was used to reliably detect illnesses in the system, which captures abnormalities in the leaves and colour changes in the leaves. Gathered a database of leaf pictures from several plant species and transferred learning to extract their characteristics. In the subsequent steps, several machine learning extraction techniques were used on the feature learning. The collected characteristics were then fed into several machine learning algorithms, and a final model with an accuracy of 94% was produced. Despite the development of several machine learning techniques to improve the overall effectiveness of plant/crop disease analysis, several aspects include Variations in disease prevalence and how well-illuminated crop photos also impact detection rates. Deep learning has a clear benefit over machine learning. Raw data may be used to apply approaches in several different file types.