Methods and apparatuses for marking target in 3D image

1. A method for marking a target in a 3-Dimensional (3D) image in a Computed Tomography (CT) system for security inspection, the method comprising:

acquiring CT image data of a scene;

rendering a 3D image of the scene using ray casting based on the CT image data;

removing a transparent region from the 3D image based on a fixed 2-Dimensional (2D) transfer function in which a fixed transparency mapping relation between a transparency of the region as well as a gray level of a region the target located in and material of the target is defined; and

marking the target in the 3D image with the transparent region being removed,

wherein said removing a transparent region further comprises:

building an octree for describing a 3D space data structure of the CT image data;

traversing the octree to determine whether each node of the octree is transparent based on a minimal gray level value, a maximal gray level value, a minimal material value, and a maximal material value of the current node by using the transparency mapping relation; and

removing each transparent node determined;

wherein the transparency mapping relation is preset by dividing a grid in the octree at a first granularity in a gray level direction and dividing the grid at a second granularity in a material direction, the first granularity is coarser than the second granularity.

2. The method according to claim 1 , further comprising after removing the transparent region:

generating an enclosure box that encloses a valid region, and generating a valid-region axial enclosure box based on the enclosure box.

3. The method according to claim 2 , wherein the valid-region axial enclosure box is used as a start position for marking the target.

4. The method according to claim 1 , wherein said removing a transparent region further comprises:

generating a lookup table of transparency maxima independent of the building of the octree.

5. The method according to claim 1 , wherein said building an octree comprises:

setting a height of the octree, and allocating the CT image data on a level basis to each node of the octree based on the height.

6. The method according to claim 1 , wherein the transparency mapping relation is preset as a lookup table for transparency maxima.

7. An apparatus for marking a target in a 3-Dimensional (3D) image in a Computed Tomography (CT) system for security inspection, the apparatus comprising a processor configured to:

acquire CT image data of a scene;

render a 3D image of the scene using ray casting based on the CT image data;

remove a transparent region from the 3D image based on a fixed 2-Dimensional (2D) transfer function in which a fixed transparency mapping relation between a transparency of the region as well as a gray level of a region the target located in and material of the target is defined; and

mark the target in the 3D image;

wherein the processor is further configured to:

build octree for describing a 3D space data structure of the CT image data;

traverse the octree to determine whether each node of the octree is transparent based on a minimal gray level value, a maximal gray level value, a minimal material value, and a maximal material value of the current node by using the transparency mapping relation; and

remove each transparent node determined;

wherein the transparency mapping relation is preset by dividing a grid in the octree at a first granularity in a gray level direction and dividing the grid at a second granularity in a material direction, wherein the first granularity is coarser than the second granularity.