Method of extracting temporal and spatial combination markers for morphological image division

• US 6,728,427 B1
• Filed: 08/02/2000
• Issued: 04/27/2004
• Est. Priority Date: 10/16/1997
• Status: Expired due to Fees

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Lee, Myoung Ho, Kim, Mun Churl, Ahn, Chie Teuk, Choi, Jae Gark
• Primary Examiner(s): Johnson, Timothy M.
• Assistant Examiner(s): Kassa, Yosef
• Application Number:
  US09/529,609
• Time in Patent Office: 1,364 Days
• Field of Search: 382/164, 382/165, 382/190, 382/205, 382/218, 382/251, 382/253, 382/260, 382/266, 382/274, 382/282, 382/283, 382/284, 382/295, 382/307, 382/308, 382/180, 382/107, 382/171, 382/177, 382/236, 358/452, 358/453, 358/450, 358/3.27, 348/14.15, 348/403, 348/405, 348/392, 348/513, 348/131, 348/132
• US Class Current: 382/308
• CPC Class Codes: H04N 19/20 using video object coding
• CPC Class Codes: H04N 19/543 using regions

Abstract

The present invention provides the method of extracting temporal and spatial combination markers including the steps of: (1) simplifying inputted current image frames, quantizing the simplified images, and then extracting spatial markers from the simplified and quantized image frames using spatial information; (2) extracting temporal markers from inputted current image frames and previous image frames using temporal information, and (3) combining the extracted spatial marker and the extracted temporal marker to extract temporal and spatial markers.

Claims

1. A method of extracting temporal and spatial combination markers applicable to a morphological image division system, comprising:

(a) simplifying inputted current image frames to divide them using filters while preserving the boundary of regions;

(b) quantizing gray levels of the brightness value of the simplified images so as to facilitate spatial marker extraction;

(c) extracting the regions the brightness value of which is uniform as spatial markers from the simplified and quantized image frames;

(d) extracting temporal markers from the inputted image frames and previous image frames using temporal information; and

(e) combining the extracted spatial marker and the extracted temporal marker to extract temporal and spatial markers.

2. The method of extracting temporal and spatial combination markers as claimed in claim 1, the step (c) comprising:

(c-1) labelling flat connected regions the brightness of which is uniform from the simplified and quantized images as a single region;

(c-2) defining, among the labelled regions, the regions the size of which is less than a given critical value as uncertain areas; and

(c-3) extracting, among the labelled regions, the regions the size of which is larger than the critical value as spatial markers.

3. The method of extracting temporal and spatial combination markers as claimed in claim 1, the step (d) comprising:

(d-1) separating inputted current image frames and previous image frames into temporally changed regions and unchanged regions using a change region detector;

(d-2) removing uncovered regions from the changed regions to divide them into moving object regions and static background regions; and

(d-3) shrinking the divided moving object regions and static background regions, respectively, forming uncertain regions between the moving object regions and the static background regions, and then extracting temporal markers including the shrunken moving object regions, the shrunken static background regions and the uncertain regions.

4. The method of extracting temporal and spatial combination markers as claimed in claim 3, the step (c) comprising:

(e-1) extracting independent moving object regions of temporal marker images contained in respective spatial markers with respect to respective spatial markers and then extracting, among them, the regions the size of which is larger than a given critical value as temporal and spatial combination markers; and

(e-2) extracting independent static background regions of temporal marker images contained in respective spatial markers and then extracting, among them, the regions the size of which is larger than the critical value as temporal and spatial combination markers.