Data fusion processing to identify obscured objects

A method for imaging obscured objects in a scene, the method comprising:

• processing three dimensional (3D) imaging data of the scene from a lidar device;
• processing 2D image data of the scene from a two dimensional (2D) image source;
• detecting an object in the scene based on the 3D imaging data;
• modeling a cloud of distributed scatterers obscuring the object in the scene based on the 3D imaging data;
• registering lidar returns of the 3D imaging data to pixels of the 2D image data;
• characterizing a material of the cloud of distributed scatterers based on the model;
• and
• altering the pixels of the 2D image data based on the characterization to reveal the object in the scene.

A non-transitory computer readable medium comprising instruction that, when executed by a processor, direct to the processor to image obscured objects in a scene, the instructions further directing the processor to:

• process three dimensional (3D) imaging data of the scene from a lidar device;
• process 2D image data of the scene from a two dimensional (2D) image source;
• detect an object in the scene based on the 3D imaging data;
• model a cloud of distributed scatterers obscuring the object in the scene based on the 3D imaging data;
• register lidar returns of the 3D imaging data to pixels of the 2D image data;
• characterize a material of the cloud of distributed scatterers based on the model;
• and
• alter the pixels of the 2D image data based on the characterization to reveal the object in the scene.