Laser beam welding is popular in microsystems technology by its specific parameters. Mostly, it is categorized by momentary nature and really short period of the method. The laser beam is concentrated on the materials to be welded and therefore, the process is usually autogenous, does not require any additional filler material. Due to, laser welding could be a fast-growing application area for industrial lasers. Many authors conducted experiments on measurement of penetration depth using large thickness of sheet by varying process parameters and developed thermal model referring to temperature distribution within work piece. The present work selected less than 1 mm stainless steel sheet and trials taken by varying combination of power, speed and sheet thickness to predict weld bead geometry and validated by image processing technique. Furthermore, the experimental setup is developed to measure online bottom surface temperature of work piece during welding by temperature data logger. Finally, found that the net measured temperature by data logger has shown sensitive indicator of penetration depth and measured value of depth of penetration by experimentation, optimization and image processing technique satisfactorily.