This work addresses the problem of angles-only target tracking, which is a highly complicated nonlinear estimation problem where a moving target has to be tracked by a moving observer. The absence of reliable information about the unknown states to be estimated and the presence of various uncertainties, render this problem as one of the most complex problems in the estimation literature. The only available information is the angle measurements that are often corrupted with noise. The conventional way is to model these noises as random variables following a Gaussian density. However, it is very often observed that the angle measurements are impulsive with large outliers. The other uncertainties are the ambiguity about the initial range and speed of the target. Thus a robust unified estimation framework is proposed by developing a parameterized maximum correntropy algorithm, using three different kernel functions namely q-Rényi, Gaussian, and Cauchy kernel functions. Then a performance evaluation is carried out by comparing the conventional new sigma point Kalman filter (NSKF) with its parametrized versions as well as the proposed range and speed-range parametrized maximum correntropy filters. The enhanced estimation accuracy of the developed algorithms is demonstrated through simulations and is observed to incur the least estimation error.