Kidney stones are unwanted waste products that are deposited in the kidneys, causing complications of urinary tract infection, calyces dilatation, and irreversible renal failure. In some cases, if stones block the urinary tract of both kidneys at the same time, the patient will have urine retention, which can be life-threatening. Therefore, early detection and treatment of kidney stones is essential. In this work, we propose a method to detect and localize kidney stones using deep learning networks such as NASNetLarge, EfficientNetB7, EfficientNetV2L, and Vision Transformer. YoloV7 is proposed to assist in kidney localization. Research results show that the proposed method with Vision Transformer technique achieves a high accuracy of 99%.