Method and apparatus for LiDAR point cloud coding

1 . A method performed by a point cloud decoding device for decoding a point cloud, the method comprising:

decoding a quantized residual block and a quantization parameter from a bitstream;

reconstructing a residual block by dequantizing the quantized residual block by using the quantization parameter;

generating a prediction block of a coding block by predicting the coding block by using stored reconstructed points;

generating reconstructed points within a frame of the point cloud by reconstructing the coding block by adding the residual block and the prediction block;

storing the reconstructed points in a buffer; and

converting a coordinate system of the reconstructed points stored in the buffer,

wherein the frame includes a set of points obtained by LiDAR, wherein the frame has a vertical length that is equivalent to a certain number of lasers and a horizontal length that is an integer value generated by dividing a total rotation angle of each laser by a rotating angular speed of each laser, and

wherein the points each have N attribute values as attribute information, where N is a natural number equal to or greater than 1.

2 . The method of claim 1 , further comprising:

decoding information for a residual coordinate system conversion from the bitstream; and

inversely converting a coordinate system of geometric information of the residual block by using the information for the residual coordinate system conversion.

3 . The method of claim 1 , wherein converting the coordinate system includes, with respect to the point cloud including the reconstructed points stored in the buffer, converting a coordinate system of geometric information from an internal coordinate system of the point cloud decoding device to a world coordinate system.

4 . The method of claim 1 , wherein:

the coding block includes one column in the frame; and

decoding the quantized residual block and the quantization parameter includes decoding the frame sequentially from a leftmost column toward a right side.

5 . The method of claim 1 , wherein:

the coding block includes at least one or more columns in the frame; and

decoding the quantized residual block and the quantization parameter includes decoding a flag indicating a presence or an absence of points in the coding block, and generating the quantized residual block based on the flag.

6 . The method of claim 5 , wherein decoding the quantized residual block and the quantization parameter includes generating the quantized residual block by decoding the N attribute values sequentially according to an attribute index with respect to the coding block.

7 . The method of claim 5 , wherein decoding the quantized residual block and the quantization parameter includes:

decoding, from the bitstream, an attribute coding order list having a length of N or less; and

generating the quantized residual block by decoding the N attribute values according to the attribute coding order list.

8 . The method of claim 7 , wherein generating the prediction block includes, when decoding the N attribute values according to the attribute coding order list with respect to the points in the coding block, predicting a current attribute value for each point by using a previously reconstructed different attribute value.

9 . The method of claim 5 , wherein decoding the quantized residual block and the quantization parameter includes, with respect to the points in the coding block, decoding coordinates of each point in the coding block and decoding the N attribute values of each of the points to generate the quantized residual block.

10 . The method of claim 5 , wherein generating the prediction block includes:

decoding a flag indicating whether a previously reconstructed block is to be used; and

when the flag is true, utilizing attribute values of the previously reconstructed block as predicted attribute values of the coding block.

11 . A method performed by a point cloud encoding device for encoding a point cloud, the method comprising:

obtaining a frame of the point cloud;

converting a coordinate system of the frame into a coordinate system internal to the point cloud encoding device;

obtaining a coding block divided from the frame;

generating a prediction block of the coding block by predicting the coding block by using stored reconstructed points;

subtracting the prediction block from the coding block to generate a residual block;

determining a quantization parameter;

quantizing the residual block by using the quantization parameter; and

encoding a quantized residual block and the quantization parameter,

wherein the frame includes a set of points obtained by LiDAR, wherein the frame has a vertical length that is equivalent to a certain number of lasers and a horizontal length that is an integer value generated by dividing a total rotation angle of each laser by a rotating angular speed of each laser, and

wherein the points each have N attribute values as attribute information, where N is a natural number equal to or greater than 1.

12 . The method of claim 11 , further comprising:

determining information for a residual coordinate system conversion;

converting a coordinate system of geometric information with respect to the residual block by using the information for the residual coordinate system conversion; and

encoding the information for the residual coordinate system conversion.

13 . The method of claim 12 , further comprising:

generating a reconstructed residual block by dequantizing the quantized residual block with the quantization parameter;

inversely converting the coordinate system of geometric information with respect to the reconstructed residual block by using the information for the residual coordinate system conversion;

generating reconstructed points of the point cloud by reconstructing the coding block by adding the reconstructed residual block and the prediction block; and

storing the reconstructed points in a buffer.

14 . A non-transitory computer-readable recording medium storing a bitstream generated by a point cloud encoding method, wherein the point cloud encoding method comprises:

obtaining a frame of a point cloud;

converting a coordinate system of the frame into a coordinate system internal to a point cloud encoding device;

obtaining a coding block divided from the frame;

generating a prediction block of the coding block by predicting the coding block by using stored reconstructed points;

subtracting the prediction block from the coding block to generate a residual block;

determining a quantization parameter;

quantizing the residual block by using the quantization parameter; and

encoding a quantized residual block and the quantization parameter,

wherein the frame includes a set of points obtained by LiDAR, wherein the frame has a vertical length that is equivalent to a certain number of lasers and a horizontal length that is an integer value generated by dividing a total rotation angle of each laser by a rotating angular speed of each laser, and

wherein the points each have N attribute values as attribute information, where N is a natural number equal to or greater than 1.