Fruit grading plays a vital role in the agricultural supply chain by ensuring consistent quality, higher market value, and improved consumer satisfaction. Traditional manual sorting and grading methods are time-consuming, labor-intensive, and highly dependent on human judgment, which often leads to inconsistency and errors in large-scale operations. To overcome these limitations, this project proposes an automatic fruit sorting and grading system that integrates computer vision, ultrasonic sensing, and neural network–based classification techniques. The system acquires real-time images of individual fruits using a camera, from which key visual features such as color, size, shape, and surface defects are extracted through image processing methods. An ultrasonic sensor is employed to estimate the physical dimensions and distance of the fruit, thereby improving the accuracy and consistency of size measurement. The extracted features are then provided as input to a trained neural network model, which classifies the fruits into predefined quality grades. The proposed system minimizes human intervention and ensures uniform, repeatable, and objective fruit quality assessment. Experimental results demonstrate improved classification accuracy and a significant increase in processing speed compared to conventional manual inspection methods. Owing to its scalability, cost-effectiveness, and reliability, the proposed approach is well suited for modern automated agriculture and post-harvest quality management applications.