Enhancing low quality images of naturally illuminated scenes

• US 20040184667A1
• Filed: 03/19/2003
• Published: 09/23/2004
• Est. Priority Date: 03/19/2003
• Status: Active Grant

• Current Assignee: Mitsubishi Electric Research Laboratories, Inc. (Mitsubishi Electric Corporation)
• Original Assignee: Mitsubishi Electric Research Laboratories, Inc. (Mitsubishi Electric Corporation)
• Inventors: Raskar, Ramesh, Ilie, Adrian, Yu, Jingi
• Granted Patent: US 7,103,227 B2
• Time in Patent Office: Days
• US Class Current: 382/254
• CPC Class Codes: G06T 11/00 2D [Two Dimensional] image generation
• CPC Class Codes: G06T 15/02 Non-photorealistic rendering
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/174 involving the use of two or more images

Abstract

A method enhances images of a naturally illuminated scene. First, a set of images I_i(x, y) is acquired of a scene. Each image is acquired under a different uncontrolled illumination. For each image I_i(x, y), intensity gradients ∇I_i(x, y) are determined, and each image is weighted with weights according to the intensity gradients. The weighted images are then combined to construct an enhanced image I′.

Claims

1. A method for enhancing images of a naturally illuminated scene, comprising:

acquiring a plurality of images I_i(x, y) of a scene, each image having a different uncontrolled illumination;

determining, for each image I_i(x, y), intensity gradients ∇

I_i(x, y);

weighting each image with weights according to the intensity gradients; and

combining the weighted images to construct an enhanced image I′

.

2. The method of claim 1 further comprising:

determining, for each image I_i(x, y) intensity gradients ∇

I_i(x, y);

determine, for each image I_i, weight W_i(x, y);

determine a gradient field G(x, y) from a weighted sum of the weights multiplied by the intensity gradients according to G(x, y)=Σ

_iW_iG_i(x, y);

construct an enhanced image I′

from the gradient field; and

normalizing pixel intensities in the enhanced image.

3. The method of claim 1 wherein the scene is static.

4. The method of claim 1 wherein the scene is dynamic.

5. The method of claim 1 wherein intensities and directions of the illumination are uncontrolled.

6. The method of claim 1 wherein the weights are based on spatial intensity gradients.

7. The method of claim 6 wherein the weights are based on temporal intensity gradients.

8. The method of claim 1 wherein each weight is either zero or one.

9. The method of claim 1 wherein the weighs are normalized in a range zero to one.