Stereo camera and imaging system

1. An imaging system, comprising:

an imaging sensor capable of operation in the visible and non-visible spectra;

a stereo camera comprising first and second apertures separated by an interocular distance and configured to focus first and second images from first and second mirrors onto a third mirror that reflects first and second light beams side by side to form a composite image on the imaging sensor;

a digital signal compression apparatus configured to encode the composite image into compressed data; and

a network connecting the stereo camera to at least one image processor and transmitting the compressed data from the stereo camera to the at least one image processor,

wherein the at least one image processor is configured to

receive the compressed data via the network,

decode the compressed data to extract first and second frames of image data from the stereo camera,

construct volumetric image data based on binocular disparity between the first and second frames,

identify at least one candidate moving object, and

clip the volumetric image data around the at least one candidate moving object to create a new image which only contains the at least one candidate moving object in a 3D spatially constrained region.

2. The imaging system of claim 1 , wherein the stereo camera comprises a mono camera including a mono surveillance camera lens and a mono surveillance camera imaging sensor as said imaging sensor and a first housing for the mono camera,

wherein the stereo camera has a second housing, and

wherein the second housing is mounted to the first housing to allow conversion of the mono camera to the stereo camera.

3. The imaging system of claim 1 , wherein the at least one image processor is configured to run a disparity matching algorithm that compares the first and second frames in determining the volumetric image data.

4. The imaging system of claim 3 , wherein the at least one processor is configured to identify the at least one candidate moving object in at least one of the first frame and the second frame, to determine a depth position of the at least one candidate moving object based on the volumetric image data and to clip the volumetric image data about the at least one candidate moving object in three dimensions in the X/Y/Z planes.

5. The imaging system of claim 4 , further comprising a video analytics processing engine configured to detect objects based on the clipped volumetric image data.

6. The imaging system of claim 4 , wherein the at least one processor is configured to identify the at least one candidate moving object based on background model subtraction.

7. The imaging system of claim 4 , wherein the at least one processor is configured to build a predetermined size three-dimensional box from a ground plane corresponding to the depth position and to clip the volumetric image data based on the predetermined size three-dimensional box.

8. The imaging system of claim 7 , wherein the predetermined size is depth adaptive.

9. The imaging system of claim 3 , further comprising:

a video analytics processing engine configured to detect at least one object based on the volumetric image data.

10. The imaging system of claim 9 , wherein the video analytics processing engine is configured to exclude at least one object in the volumetric image data from further processing based on at least one of size in at least one dimension, height from ground plane and speed of movement.

11. The imaging system of claim 10 , wherein the further processing is by an object classifier.

12. An imaging system, comprising:

an imaging sensor capable of operation in the visible and non-visible spectra;

a stereo camera comprising first and second apertures separated by an interocular distance and configured to focus first and second images from first and second mirrors onto a third mirror that reflects first and second light beams side by side to form a composite image on the imaging sensor; and

an image processor, configured to

receive first and second frames of image data from the stereo camera,

construct volumetric image data based on binocular disparity between the first and second frames,

run a disparity matching algorithm that compares the first and second frames in determining the volumetric image data,

identify at least one candidate moving object in at least one of the first frame and the second frame,

determine a depth position of the at least one candidate moving object based on the volumetric image data,

determine a 3D bounding box about each at least one moving object,

determine a centroid for each 3D bounding box,

determine a depth position of the centroid based on the volumetric image data, and

clip the volumetric image data about the at least one candidate moving object along the associated 3D bounding box in the X/Y/Z planes.

13. An imaging system, comprising:

a plurality of stereo cameras each comprising

an imaging sensor capable of operation in the visible and non-visible spectra, and

an optical apparatus comprising first and second apertures separated by an interocular distance and configured to focus first and second images from first and second mirrors onto a third mirror that reflects first and second light beams side by side to form a composite image on the imaging sensor; and

a network connecting the plurality of stereo cameras to an image processor, each camera transmitting imaging data over the network;

wherein the image processor is configured

to receive the imaging data from the network and extract first and second frames of image data corresponding to the side by side composite images of one of the plurality of stereo cameras, and construct volumetric image data based on binocular disparity between the first and second frames, and

to identify at least one candidate moving object, and

to clip the volumetric image data around the at least one candidate moving object to create a new image which only contains the at least one candidate moving object in a 3D spatially constrained region.