A Hybrid Approach for Indian Coin Classification Using CNN and KNN

The summary emphasizes the successful deployment of a strong coin class model tailored specially for Indian coins. The model has tremendous overall performance measures, including high precision, keep in mind, and F1 rankings throughout a couple of denominations. For example, the 1 rupee coin had a precision, don’t forget, and F1-rating of 98.59, while the 2 rupee coin marginally outperformed it with values of 98.62. The 10 and 20 rupee coins likewise had top metrics, scoring 98.64 and 98.60, respectively. Although the 5 rupee coin had barely decreased outcomes (accuracy, bear in mind, and F1-rating of 95.65), it demonstrated amazing categorization abilities. With 10,870 pics as a dataset, the fashions used yielded a normal accuracy of 98.62%. The model’s performance is in addition supported by way of balanced magnificence distributions, as indicated through the support values: 2,130 for the 1 Rupee, 2,180 for the 2 Rupee, 2,220 for the 5 Rupee, 2,220 for the 10 Rupee, and 2,140 for the 20 Rupee cash, all of which contribute to a steady accuracy of 0.99 (99%) in all training. This consistency no longer handiest will increase the version’s trustworthiness however also helps its suitability to be used in automatic coin reputation structures. The version’s excessive type accuracy suggests that it’s miles well-appropriate to supporting users in reliably figuring out diverse Indian coin denominations, making it a vital device for monetary transactions and automated merchandising systems. Overall, the consequences display the cautioned model’s usefulness and performance in fixing the issues of coin categorization in real-world circumstances.