Fruits are the most popular demanding product in the market, so the quality of those products should be very good. Automation in the field of fruit classification can bring great continence to fruit mongers. Creating large data banks to store information about diseases and use them with techniques such as deep learning will be helpful for economic development. This paper will be helpful to the agricultural industry. However, some pears and apple varieties have many similarities, and these sorts of fruit are typically popular with consumers. These things make the task of identifying rotten fruits more difficult. To solve this issue, this research proposes a convolutional neural network-based deep learning method for fruit classification. The dataset used contains images of different fruits. Applying a convolutional neural network, we will categorize the images of fruit into fresh and rotten categories. Several tests have been conducted on those images to get an efficiency of 94.92%.