Machine Learning algorithms are an important aspect of computer vision in the field of robotics. And with the exponential and advent growth of Deep Learning research along with advanced neural networks, Computer Vision has strengthened its role and contribution in crafting the abilities of autonomous robots, allowing them to be capable of performing high-end tasks, including but not limited to object detection, navigation, manipulation and other important aspects that autonomous robots may be required for. This review paper explored several machine learning approaches like Supervised Learning, Unsupervised learning and reinforcement Learning, discussing the contributions and applications of these approaches used in the field of computer vision for robotics. This paper also discusses several key architectures, such as Convolutional Neural Networks (CNNs) and state-of-the-art models such as Faster R-CNN and YOLO, few of which are discussed in detail. Several challenges and limitations linked to real-time processing, data quality and their ethical considerations are also discussed in detail. These discussions are supported by through case studies and performance evaluations, which aim to provide further insights into current and future advancements along with identifying future directions in the integration of machine learning and computer vision for robotics.