We present a range-gating delayed detection super-resolution imaging lidar with high accuracy based on the signal intensities of three consecutive delay samples. The system combines the range and signal intensity information from multi-pulse detections to calculate the pulse peak position under the assumption of a Gaussian pulse shape. Experimental results indicate that the proposed algorithm effectively calculates pulse peak position and exhibits excellent accuracy with super-resolution. Accuracy analysis shows that accuracy is best improved by enhancing signal-to-noise ratio, strategically selecting samples, reducing pulse width, and appropriately choosing the delayed periods between samples.