Maximum Likelihood and Kernel Estimate Methods for Line Transect Density Estimation

The probability density function at perpendicular distance zero from data obtained by the line transects method. A nonparametric kernel method produces asymptotic unbiased estimator for f(0) provided that the detection model has a shoulder at distance zero, that is, f (0) = 0. If the shoulder condition seems to be valid using as reference the half normal density, while if the shoulder condition does not seem to be valid, use the exponential density as a reference. The results demonstrate the superiority of the better estimator in most cases considered because without the shoulder condition, kernel estimate is not as good as maximum likelihood, and maximum likelihood estimate is quite effective. But, kernel estimate after improving is better than maximum likelihood estimate for line transect sampling.