Iris recognition continues to be the pinnacle of biometric security systems, but it is still facing challenges and issues with iris images with noisy factors or commonly known as non-ideal iris images. In recent years it has gained significant attention to the development of a robust and dependable iris recognition system. However, the accuracy of iris segmentation is often affected by off-angle and non-ideal iris images, remains a challenging task. In this paper, we are going to review the enhanced segmentation method to deal with off-angle and non-ideal iris images. The development is crucial to overcome the challenges faced in accurately identifying both off-angle and non-ideal iris images, which is of paramount importance for the security and identification applications that heavily rely on iris recognition technology. Even so, it is still facing major issues such as occlusions, gaze/off angle, specular reflection, presence of contact lenses and eyeglasses, motion blur etc. The review method incorporates enhancements and deep learning method strategies that improve the segmentation accuracy, particularly on off-angle and non-ideal iris images to enhance the overall performance of iris recognition. We also review various public datasets that were used to train and test in literature such as CASIA, and UBIRIS along with the types of non-ideal iris images.