Laser-induced　coloration　on　a　metallic　surface　has　been　of　interest　to　many　application　arweas.　However　laser　machining　of　metals　involves　many　complex　problems　including　nonlinear　unstable　coupled　with　multiple　factors.　Therefore　there　are　still　some　significant　challenges　in　the　precise　control　of　color　creation.　Here　we　explored　the　process　of　the　laser-induced　coloration　and　find　the　connection　between　surface　colors　and　processing　parameters.　The　Response　Surface　Methodology　(RSM)　based　experimental　design　was　adopted　to　explore　the　influence　of　the　single　processing　parameter　and　the　interaction　between　parameters　on　color　changes　of　titanium.　The　results　showed　that　the　scanning　speed　laser　power　repetition　rate　and　hatch　distance　had　significant　effects　on　color　changes　of　titanium.　Then　we　demonstrated　that　using　artificial　neural　network　(ANN)　is　an　effective　solution　of　nonlinear　problems　in　laser-induced　coloration　which　can　match　the　processing　parameters　and　the　L*a*b*　color　values　on　titanium　surface　precisely　with　limited　experiments.　Finally　we　successfully　used　the　processing　parameters　estimated　by　ANN　model　to　create　unique　art　painting　on　titanium　with　nanosecond　pulsed　laser.　This　work　can　provide　a　potential　method　to　solve　the　problem　in　the　color　consistency　and　open　a　new　perspective　in　industrial　application　of　laser-induced　coloration　technology.