APPARATUS AND METHOD FOR CLASSIFYING OBJECTS IN COMPUTED TOMOGRAPHY DATA USING DENSITY DEPENDENT MASS THRESHOLDS

A method and apparatus for detecting objects in computed tomography (CT) data are disclosed. Sheet-shaped objects such as sheet explosives can be detected by analyzing a neighborhood of voxels surrounding a test voxel. If the density of the test voxel is sufficiently different from the mean density of the neighboring voxels, then it is concluded that the test voxel is associated with a sheet object. Sheet objects can also be detected by eroding the CT data so as to eliminate voxels associated with thin objects. Remaining objects are then subtracted from the original data, leaving only thin sheet-shaped objects. Erosion of the data can be performed by identifying a neighborhood of voxels surrounding a voxel of interest. If the number of voxels having densities below a predetermined threshold exceeds a predetermined number, then it is assumed that the test voxel is a surface voxel and is removed from the object. A connectivity process can be applied to voxels to combine them into objects after sheets are detected to prevent sheets from being inadvertently removed from the data by erosion. A dilation function can then be performed on the eroded object to replace surface voxels removed by erosion. A corrected mass using the mean eroded density of the object can be computed and compared to mass thresholds to classify the object as to whether it poses a threat. Multiple mass thresholds can be used, each of which is associated with a particular density range based on the density of an expected threat object. Bulk objects can be detected by a modified morphological connected components labeling (CCL) approach which performs a series of erosion and dilation steps to separate adjacent objects in the data such that they can be individually labeled and analyzed. A merging process can be used to reconnect related items, such as multiple sticks, that were separated during an erosion step. The merging process allows multiple objects that would individually pass as non-threat items to be combined into a single item that is correctly classified as a threat. The system can also identify objects that contain liquids, if desired. The process of the invention can be carried out in multiple stages. The overall system performance, including overall object detection rate and false alarm rate, can be adjusted by adjusting individual object detection rates and/or false alarm rates.