Attribute-based point cloud strip division method
View Patent ↗An attribute-based point cloud strip division method. The method comprises: first, performing spatial division of a certain depth on a point cloud to obtain a plurality of local point clouds; and then, sorting the attribute values in the local point clouds, and on the basis of such, further performing point cloud strip division so as to obtain point cloud strips that have low geometric overhead and a uniform number of points. By means of comprehensively using the spatial position and attribute information of the point clouds, the points having similar attributes and related spatial positions are gathered as much as possible in one strip during strip division, which is convenient for making full use of the redundancy of the attribute information between adjacent points, and improving the performance of point cloud attribute compression. At the same time, independent coding between strips supports random access, improves coding efficiency, and prevents the accumulation and spread of coding errors, thus enhancing the fault tolerance of the system.
1. A strip division method for transforming a three dimensional object into digital form, wherein firstly spatial partitioning is performed on a point cloud at a certain depth to obtain local point clouds; then attribute types in each point cloud of the local point clouds are sorted and further point cloud partitioning is performed to obtain a point cloud strip with low geometric overhead and uniform points; the method comprises the following steps of:
1) preliminarily partitioning a point cloud to obtain the local point clouds by splitting geometric space of the point cloud, and then obtaining one group of local point clouds with sequence by sorting;
2) strip division of the local point clouds based on sorting by attribute type:
wherein the sorting by attribute type comprises iterating all local point clouds, and sorting attribute types in each point cloud of the local point clouds,
wherein the attribute types comprise a color value and a resolution ratio value;
if the attribute type of the point cloud is the color value, then the color values are sorted in ascending order, and a current local point cloud is divided into uniform points according to a setting of a number of a strip;
if the attribute type of the point cloud is the resolution ratio value, then the resolution ratio values are sorted in ascending order, and the current local point cloud is divided into uniform points according to the setting of the number of the strip; and
if the point cloud contains attribute types of both the color value and the resolution ratio value at the same time, then color values are taken as representatives to carry out ascending order sorting;
the above steps making it possible that the point cloud strip with uniform points and close attribute types of the three dimensional object is obtained.
2. The method according to claim 1 , characterized in that in step 1), the geometric space of the point cloud is preliminarily divided by adopting a binary tree: there being N points in total of the point cloud to be processed, division depth being set by the binary tree as d, and 2 d local point clouds being obtained after the point cloud is divided d times; then, all local point clouds being numbered b 1 , b 2 , . . . , b i , . . . , b 2 d according to an order of breadth traversal.
3. The method according to claim 2 , characterized in that the step 1) further comprises: performing spatial division on the point cloud according to geometric information, selecting a coordinate axis with the largest distribution variance in point cloud position coordinate as a division axis each time, selecting a point with a coordinate size being a median value as a division point, performing iterative division until a set depth of the binary tree is reached, and obtaining local point cloud with almost equal points after division.
4. The method according to claim 1 , characterized in that in step 2), ascending order is performed on a brightness component of a color; assuming that color values in the local point cloud b (I) are R(n), G(n), B(n), the calculation formula of the brightness component is as follows:
Y ( n )=round(0.2126* R ( n )+0.7152* G ( n )+0.0722* B ( n ))
5. The method according to claim 1 , characterized in that details of the division of the local point cloud in step 2) are as follows: assuming that strip division number of the current point cloud is Num, 2 d local point clouds can be obtained through step 1), and in order to ensure that the points in each local point cloud are uniform, the strip number num of each local point cloud after re-division is as follows:
num =ceil( Num/ 2 d )