<p>In a typical human kinematic analysis, kinematic parameters change dynamically as a person walks. Various methods are available to evaluate kinematic gait parameters, such as joint angles. Optical motion capture systems is one such method, which use markers affixed on the human body to evaluate joint angles, are commonly employed. The movement of these markers is captured by video cameras, and the footage is analysed using specialized software. The specialised softwares are intensive programmed algorithm, which detects, tracks and calculate the gait parameters. Use of such systems are expensive and require skilled operators. The present study aims to outline a novel, cost effective approach for capturing video and analyzing data using a Python algorithm. Active light-emitting diode (LED) markers are used as joint markers, and the Euclidean Distance Tracker (EDT) algorithm is employed for marker detection and tracking that also calculates the knee flexion–extension angle for one gait cycle. The root mean square error (RMSE) is employed to compare the accuracy of calculated knee flexion–extension angle with reference data, obtained from sensor-based gait measurement system. The EDT algorithm have been tested with different factors and key input factors are, camera settings (ISO and shutter speed) and ambient settings (marker colour, background colour, and clothing colour). The output responses analysed include bounding box area (BBA), peak knee flexion angle and file size. Interaction plots between response variables and input factors were used to identify the best settings. The results demonstrate that adjusting the background to blue colour produces lower RMSE of 8.74° when compared to the reference data. Additionally, green LED markers yield the best overall results. Marker-based 2D gait analysis calculations can be most accurately performed with these optimized camera settings and ambient settings.</p>