Photographs of five distinct rice varieties were classified using the new diffusion convolutional neural network (DCNN) technique to create a rice detection system. In this study, we employed a 48-megapixel iPhone 16 Plus camera, which utilizes sensor technology to take 1000 sample photos under various lighting conditions, such as day and night. Regarding the technical approach, a DCNN based on deep learning was used to categorize rice. By calculating the indication of each performance metric, the examined classes generated an overall accuracy of 99.0% using the dataset for training, testing, and validation. In addition, six supervised learning and two deep learning algorithms were tested on these rice varieties and the results were compared. Finally, the practicality of the DCNN tests employing a larger input publicity dataset was assessed, along with their accuracy, loss, and training time. Statistical analysis and comparison showed that our technique achieved a 99% classification rate. They also explain the benefits of DCNN technology compared it with other models, achieving higher performance for agricultural data. On the other hand, integrating GPS into rice seed classification is an actual use of sensor technology, especially in DCNN methodology related to machine learning.