Region-of-interest video quality enhancement for object recognition

1. A method of providing video-based object recognition, the method comprising:

receiving a frame of image data from an image capture device;

selecting a region-of-interest (ROI) within the received frame;

calculating a video quality metric (VQM) value for the selected ROI defining quality of the selected ROI;

selecting a video quality enhancement (VQE) function based on the VQM value calculated with respect to the ROI;

applying the selected VQE function to the selected ROI to enhance quality of the ROI;

applying the selected VQE function to the ROI with a plurality of discrete parameter values to generate a plurality of enhanced ROIs, wherein the one or more parameter values define the extent of enhancement provided to the ROI by the selected VQE function;

calculating an objective metric measuring quality of each of the plurality of enhanced ROIs;

selecting the parameter value that provided the highest quality objective metric as part of the selected VQE function; and

applying an object recognition algorithm to the enhanced ROI to identify objects within the ROI.

2. The method of claim 1 , wherein selecting the ROI includes:

applying an object detection algorithm to automatically identify objects of interest within the received image data; and

selecting the ROI based on the automatically identified objects.

3. The method of claim 1 , wherein selecting the ROI includes receiving manual input from a user identifying the ROI.

4. The method of claim 1 , wherein calculating the VQM value includes calculating a 20 VQM value selected from a group consisting of: sharpness/blurring metrics, contrast metrics, brightness metrics, noise metrics, color metrics, and combinations thereof.

5. The method of claim 1 , wherein selecting the VQE function includes adjusting parameters associated with the selected VQE function based on the calculated VQM value to provide a desired amount of enhancement to the ROI.

6. A video-based object recognition system comprising:

a video processor, configured to receive a video input, the video processor comprising—

an object detection module that identifies objects of interest within a frame of image data from the video input, wherein a region of interest (ROI) is defined to include the identified object;

a video quality metric (VQM) module that calculates one or more VQM values for the selected ROI that defines image quality of the selected ROI;

a video quality enhancement (VQE) module that selects a VQE function to apply to the defined ROI based on the one or more VQM values calculated with respect to the defined ROI to generate an enhanced ROI;

an objective quality metric module that calculates a quality value associated with application of enhancements to the selected ROI, wherein the VQE module employs an iterative approach to adjusting parameter values associated with a selected VQE function to select parameters that maximize the quality value calculated by the objective quality metric module, wherein the one or more parameter values define the extent of enhancement provided to the ROI by the selected VQE function; and

an object recognition module that recognizes objects within the defined ROI based on the enhanced ROI provided by the VQE module.

7. The video-based object recognition system of claim 6 , wherein the object detection module receives manual input from a user to define the ROI.

8. The video-based object recognition system of claim 6 , wherein the object detection module includes an algorithm for automatically analyzing the frame of image data and identifying objects of interest.

9. The video-based object recognition system of claim 6 , wherein the VQM module calculates a VQM value selected from a group consisting of: sharpness/blurring metrics, contrast metrics, brightness metrics, noise metrics, color metrics, and combinations thereof.

10. The video-based object recognition system of claim 6 , wherein the VQE module includes one or more parameters associated with each VQE function that are selectively adjusted by the VQE module based on the VQM value calculated by the VQM module, wherein the one or more parameter values define an extent of enhancements provided to the ROI by the selected VQE function.