测绘与空间地理信息
測繪與空間地理信息
측회여공간지리신식
GEOMATICS & SPATIAL INFORMATION TECHNOLOGY
2014年
4期
118-121,124
,共5页
车载激光扫描%道路标识线%Hough变换%图像分割
車載激光掃描%道路標識線%Hough變換%圖像分割
차재격광소묘%도로표식선%Hough변환%도상분할
vehicle-borne laser scanning%road mark%Hough transform%image segmentation
近年来,随着空间信息获取技术的发展,激光扫描技术在城市三维数据采集中应用越来越广泛,本文以车载激光扫描点云数据为研究对象,利用点云数据空间分布特征和反射强度信息,结合道路标线的几何特征,提出一种快速有效地从离散点云中提取道路标识线的方法。该方法首先利用车载激光点云数据中的高程信息和反射强度信息对原始点云进行滤波。然后将分割后的点云数据投影到二维平面中,利用反射强度信息和点云空间分布信息生成点云强度特征图像,利用标线规则的几何形状,对连通区域进行道路标识线的提取。最后,基于道路标识线的语义信息,利用Hough变换对检测到的标识线进行分类和连接,从而提取完整、准确的三维道路标识线点云数据。通过居民区和高速公路扫描数据处理案例,实现了高速公路虚实标识线和干扰因素较多的居民区界线的自动提取,验证了上述道路标识线提取方法的可靠性,应用效果较好。
近年來,隨著空間信息穫取技術的髮展,激光掃描技術在城市三維數據採集中應用越來越廣汎,本文以車載激光掃描點雲數據為研究對象,利用點雲數據空間分佈特徵和反射彊度信息,結閤道路標線的幾何特徵,提齣一種快速有效地從離散點雲中提取道路標識線的方法。該方法首先利用車載激光點雲數據中的高程信息和反射彊度信息對原始點雲進行濾波。然後將分割後的點雲數據投影到二維平麵中,利用反射彊度信息和點雲空間分佈信息生成點雲彊度特徵圖像,利用標線規則的幾何形狀,對連通區域進行道路標識線的提取。最後,基于道路標識線的語義信息,利用Hough變換對檢測到的標識線進行分類和連接,從而提取完整、準確的三維道路標識線點雲數據。通過居民區和高速公路掃描數據處理案例,實現瞭高速公路虛實標識線和榦擾因素較多的居民區界線的自動提取,驗證瞭上述道路標識線提取方法的可靠性,應用效果較好。
근년래,수착공간신식획취기술적발전,격광소묘기술재성시삼유수거채집중응용월래월엄범,본문이차재격광소묘점운수거위연구대상,이용점운수거공간분포특정화반사강도신식,결합도로표선적궤하특정,제출일충쾌속유효지종리산점운중제취도로표식선적방법。해방법수선이용차재격광점운수거중적고정신식화반사강도신식대원시점운진행려파。연후장분할후적점운수거투영도이유평면중,이용반사강도신식화점운공간분포신식생성점운강도특정도상,이용표선규칙적궤하형상,대련통구역진행도로표식선적제취。최후,기우도로표식선적어의신식,이용Hough변환대검측도적표식선진행분류화련접,종이제취완정、준학적삼유도로표식선점운수거。통과거민구화고속공로소묘수거처리안례,실현료고속공로허실표식선화간우인소교다적거민구계선적자동제취,험증료상술도로표식선제취방법적가고성,응용효과교호。
The car-mounted mobile laser scanning system ( mobile lidar) has become a cost-effective solution for capturing spatial data in complex urban areas at the street level quickly and accurately .In particular , mobile lidar can acquire three -dimensional (3D) dense -points, which enable easier building fa?ade reconstruction, man-made object extraction, 3D city modeling, street-scene modeling and visualization than most other methods .In most By inherent characteristics of scanning data , the rapid and effective method used to extracting road mark was put forward with the study of vehicle -borne laser scanning point cloud data .The proposed method first generates a geo -referenced feature image of point clouds followed by extracting the approximate outline of road using dis -crete discriminate analysis from the geo -referenced feature image .Second , the proposed method further segments the point clouds of road according to the reflectance strength of point clouds for extracting the point clouds of road markings .Generally, road markings show a set of predefined shapes .Hence, these predefined shapes can easily be modeled as many structures (e.g., rectangle, dotted lines).The proposed method generates a geo -referenced reflectance strength image of the point clouds of road markings , labeling the regions of road markings according to the shape and arrangement of road markings , and applies the PPHT operator for extracting road markings .In order to improve the performance of road markings extraction , the proposed method incorporates the semantic knowledge (e.g., shape, pattern) of road markings, which is particularly helpful for extracting those road markings occluded or with incomplete shapes.The above key steps were successfully integrated into the proposed method .Two different datasets were selected for checking the validity of the proposed method .The experimental results demonstrated that the proposed method shows a good performance for ex -tracting road marking from mobile lidar point clouds .Moreover , the match between the extracted road markings and the original point clouds proves that the proposed method detects and extracts road markings correctly .