In　order　to　implement　the　edge　enhancement　and　lane　line　detection　of　the　shadowy　and　strong　illumination　et　al,　an　image　processing　method　based　on　fuzzy　logic　is　proposed.　The　maximum　information　entropy　principle　is　introduced　to　obtain　the　parameter　of　the　membership　function　through　the　histogram.　The　fuzzy　logic　method　is　used　to　enhance　the　contrast　between　lane　line　pixels　and　road　pixels.　During　the　lane　detection,　the　HT　(Hough　transform)　is　employed　to　implement　the　lane　straight　line　detection　and　the　3rd　order　curve　model　is　used　to　curve　line　fitting　after　image　preprocessing　using　fuzzy　logic.　In　order　to　reduce　computational　time,　the　Kalman　filter　is　used　to　establish　the　dynamic　ROI　(region　of　interest)　and　predict　the　lane　line　parameters　for　lane　line　fitting　according　to　the　lane　parameters　of　former　image　for　real-time　detection.　By　comparison,　it　indicates　that　the　method　can　intensify　lane　lines　information　effectively　and　enhance　the　contrast　between　the　lane　line　pixels　and　road　pixels.　Experiments　results　show　that　lane　line　detection　using　fuzzy　logic　can　estimate　the　lane　line　parameters　accurately　and　detect　the　lane　lines　stably　under　different　illumination　conditions.