Method for building a three-dimensional model and apparatus thereof
A method to build a 3D model for a physical object includes the following steps. First, depth information of a plurality of images is converted into a plurality of 3D point clouds. Then, the motion parameters of the camera to gather each image relative to a previous image are determined according to the 3D point clouds. Next, a registration energy estimate value of each image's 3D point cloud, as being integrated into a previous image's 3D point cloud, is determined according to corresponding motion parameters. Then, the motion parameters are varied to minimize the estimate value. An optimal camera path is determined according to the varied motion parameters. Finally, a 3D model of the physical object is built according the optimal camera path.
1. A method to build a 3D model for a physical object, comprising:
converting depth information of a plurality of images into a plurality of 3D point clouds, wherein the images are gathered by a camera surrounding the physical object to sequentially photograph the physical object;
respectively calculating motion parameters of the camera gathering each of the images relative to gathering a previous image according to the 3D point clouds;
calculating a registration energy estimate value of the 3D point cloud for each of the images as being integrated into the 3D point cloud of the previous image according to a corresponding motion parameter;
varying the motion parameters to minimize the registration energy estimate value;
determining an optimal camera path according to the varied motion parameters; and
building a 3D model of the physical object according the optimal camera path.
2. The method of claim 1 , wherein building a 3D model of the physical object according the optimal camera path further comprises integrating the 3D point clouds according to the varied motion parameters.
3. The method of claim 1 , wherein minimizing the registration energy estimate value uses a Particle Swarm Optimization (PSO) method or a Genetic Algorithms (GA) method.
4. The method of claim 1 , further comprising converting the depth information of the plurality of images into pixel color information.
5. The method of claim 4 , further comprising calculating a pixel color registration energy estimate value of the 3D point cloud of each of the images as being integrated into the 3D point cloud of the previous image according to a corresponding motion parameter.
6. The method of claim 1 , wherein each of the motion parameters includes rotation vectors and moving vector.
7. The method of claim 1 , wherein respectively calculating motion parameters of the camera gathering the images according to the 3D point clouds further comprises:
calculating a motion parameter change of the camera gathering each of the images relative to gathering the previous image;
determining whether or not the motion parameter change is in a set range; and
varying the motion parameters when the motion parameter change is out the set range.
8. The method of claim 7 , wherein an interpolation method is used to vary the motion parameters.
9. A non-transitory computer readable storage medium having a software, wherein the software is processed by a processor to perform the method as claimed in claim 1 .
10. An apparatus to build a 3D model for a physical object, comprising:
an input for receiving depth information of a plurality of images, wherein the images are gathered by a camera surrounding the physical object to sequentially photograph the physical object;
a converter for converting the depth information into a plurality of 3D point clouds;
a motion parameter generator for generating motion parameters of the camera gathering each of the images relative to gathering a previous image according to the 3D point clouds;
a first calculator for calculating a registration energy estimate value of the 3D point cloud of each of the images as being integrated into the 3D point cloud of the previous image according to a corresponding motion parameter;
a second calculator for varying the motion parameters to minimize the registration energy estimate value;
a camera path generator for determining an optimal camera path according to the varied motion parameters; and
a 3D model generator for building a 3D model of the physical object according the optimal camera path.
11. The apparatus of claim 10 , wherein the 3D model generator integrates the 3D point clouds according to the varied motion parameters to build the 3D model of the physical object.
12. The apparatus of claim 10 , wherein the second calculator uses a Particle Swarm Optimization (PSO) method or a Genetic Algorithms (GA) method to minimize the registration energy estimate value.
13. The apparatus of claim 10 , wherein the converter further converts the depth information of the plurality of images into pixel color information.
14. The apparatus of claim 13 , wherein the first calculator calculates a pixel color registration energy estimate value of the 3D point cloud each of the images being integrated into the 3D point cloud of the previous image according to a corresponding motion parameter.
15. The apparatus of claim 10 , wherein each of the motion parameters includes rotation vectors and moving vector.
16. The apparatus of claim 10 , wherein the motion parameter generator further calculates a motion parameter change of the camera gathering each of the images relative to gathering the previous image, and varies the motion parameters when the motion parameter change is out the set range.
17. The apparatus of claim 16 , wherein an interpolation method is used to vary the motion parameters.