Method for detecting a face in a digital image

• US 6,263,113 B1
• Filed: 12/11/1998
• Issued: 07/17/2001
• Est. Priority Date: 12/11/1998
• Status: Expired due to Term

• Current Assignee: Philips Electronics North America Corporation (Koninklijke Philips N.V.)
• Original Assignee: Philips Electronics North America Corporation (Koninklijke Philips N.V.)
• Inventors: Elgammal, Ahmed, Abdel-Mottaleb, Mohamed
• Primary Examiner(s): Couso, Jose L.
• Assistant Examiner(s): Do, Anh Hong
• Application Number:
  US09/210,415
• Time in Patent Office: 949 Days
• Field of Search: 382/237, 382/118, 382/170, 382/266, 382/190, 382/308, 382/199, 382/224, 384/521, 384/522, 348/169, 348/77
• US Class Current: 382/237
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships

Abstract

In order to detect a face disposed within a digital image, the pixels of the image are grouped based on whether they are skin color. The edges of the skin colored areas are removed by eliminating pixels that have surrounding pixels with a high variance in the luminance component. The resulting connected components are classified to determine whether they could include a face. The classification includes examining: the area of the bounding box of the component, the aspect ratio, the ratio of detected skin to the area of the bounding box, the orientation of elongated objects, and the distance between the center of the bounding box and the center of mass of the component. Components which are still considered facial candidates are mapped on to a graph. The minimum spanning trees of the graphs are extracted and the corresponding components which remain are again classified for whether they could include a face. Each graph is split into two by removing the weakest edge and the corresponding components which remain are yet again classified. The graph is continually broken down until a bounding box formed around the resulting graphs is smaller than a threshold. Finally, a heuristic is performed to eliminate false positives. The heuristic compares the ratio of pixels with high variance to the total number of pixels in a face candidate component.

Claims

1. A method for detecting a face disposed within a digital image, comprising the steps of:

providing a digital image composed of a plurality of pixels;

providing a binary image from the digital image by detecting skin colored pixels;

removing pixels corresponding to edges in the luminance component of said binary image thereby producing binary image components;

mapping said binary image components into at least one graph; and

classifying said mapped binary image components as facial and non-facial types wherein the facial types serve as facial candidates,

further comprising the step of applying a heuristic, said heuristic including in the following steps;

applying a morphological closing operation on each of said facial candidates to produce at least one closed facial candidate;

determining high variance pixels in said closed facial candidate;

determining the ratio between said high variance pixels and the total number of pixels in said closed facial candidate; and

comparing said ratio to a threshold.

2. A method for detecting a face disposed within a digital image, comprising the steps of:

providing a digital image composed of a plurality of pixels;

providing a binary image from the digital image by detecting skin colored pixels;

removing pixels corresponding to edges in the luminance component of said binary image thereby producing binary image components;

mapping said binary image components into at least one graph; and

classifying said mapped binary image components as facial and non-facial types wherein the facial types serve as facial candidates

wherein said step of removing includes;

applying a mask to a plurality of said pixels including an examined pixel;

determining the variance between said examined pixel and pixels disposed within said mask; and

comparing said variance to a variance threshold wherein;

said step of removing is repeated for decreasing variance thresholds until a size or said binary image components is below a component size threshold; and

after each step of removing, each of said binary image components is classified as one of the facial type and non-facial type.

3. A method for detecting a face disposed within a digital image, comprising the steps of:

providing a digital image composed of a plurality of pixels;

providing a binary image from the digital image by detecting skin colored pixels;

removing pixels corresponding to edges in the luminance component of said binary image thereby producing binary image components;

mapping said binary image components into at least one graph; and

classifying said mapped binary image components as facial and non-facial types wherein the facial types serve as facial candidates wherein said step of removing includes;

applying a mask to a plurality of said pixels including an examined pixel;

determining the variance between said examined pixel and pixels disposed within said mask; and

comparing said variance to a variance threshold wherein;

said step of removing is repeated for decreasing variance thresholds until a size of said binary image components is below a component size threshold; and

after each step of removing, each of said binary image components is classified as one of the facial type and non-facial type,

wherein said binary image components are connected.

4. A method for detecting a face disposed within a digital image, comprising the steps of:

providing a digital image composed of a plurality of pixels;

providing a binary image from the digital image by detecting skin colored pixels;

removing pixels corresponding to edges in the luminance component of said binary image thereby producing binary image components;

mapping said binary image components into at least one graph; and

classifying said mapped binary image components as facial and non-facial types wherein the facial types serve as facial candidates,

wherein said step of mapping comprises the following steps;

representing each component as a vertex;

connecting vertices with an edge when close in space and similar in color, thereby forming said at least one graph.

5. A method for detecting a face disposed within a digital image, comprising the steps of:

providing a digital image composed of a plurality of pixels;

providing a binary image from the digital image by detecting skin colored pixels;

removing pixels corresponding to edges in the luminance component of said binary image thereby producing binary image components;

mapping said binary image components into at least one graph; and

classifying said mapped binary image components as facial and non-facial types wherein the facial types serve as facial candidates,

wherein said step of mapping comprises the following steps;

representing each component as a vertex;

connecting vertices with an edge when close in space and similar in color, thereby forming said at least one graph,

wherein each edge has an associated weight and further comprising the steps of;

extracting the minimum spanning tree of each graph;

classifying the corresponding binary image components of each graph as one of the facial type and non-facial type;

removing the edge in each graph with the greatest weight thereby forming two smaller graphs; and

repeating said step of classifying the corresponding binary image components for each of said smaller graphs until a bounding box around said smaller graphs is smaller than a graph threshold.