Process for the stabilization of the images of a scene correcting offsets in grey levels, detection of mobile objects and harmonization of two snapshot capturing apparatuses based on the stabilization of the images

• US 7,095,876 B2
• Filed: 07/31/2002
• Issued: 08/22/2006
• Est. Priority Date: 07/31/2001
• Status: Active Grant

• Current Assignee: Sagem Defense Securite (Safran SA)
• Original Assignee: Sagem SA (Safran SA)
• Inventors: Broekaert, Michel
• Primary Examiner(s): Wu, Jingge
• Assistant Examiner(s): TABATABAI, ABOLFAZL
• Application Number:
  US10/208,883
• Publication Number:
  US 20030031382A1
• Time in Patent Office: 1,483 Days
• Field of Search: 382/103, 382/107, 382/209, 382/264, 382/286, 382/291, 382/100, 382/269, 348/97, 348/135, 348/169, 348/154, 348/155, 348/208.1, 348/208.2, 348/208.4, 348/208.13, 348/699, 348/352, 348/451, 348/452, 348/700, 356/23, 356/29, 356/43, 356/149, 734/88
• US Class Current: 382/107
• CPC Class Codes: G06T 7/277 involving stochastic approaches, e.g. using Kalman filters

Abstract

A process for the electronic stabilization of the images of a scene of a snapshot capturing apparatus of an imaging system in which, in a terrestrial reference frame, the images of the scene captured by the apparatus are filtered in a low-pass filter, so as to retain only the low spatial frequencies thereof, and the optical flow equation is solved to determine the rotations to be imposed on the images in order to stabilize them with regard to the previous images is provided. The invention also relates to a process for correcting offsets in grey levels of a snapshot capturing apparatus of an imaging system whose images are stabilized according to the process of the invention, where the calculation of the offsets is deduced from the steps of the stabilization process.

Claims

1. Process for correcting offsets in grey levels of a snapshot capturing apparatus of an imaging system whose images are stabilized according to a process for the electronic stabilization of the images of a scene of a snapshot capturing apparatus of an imaging system in which, in a terrestrial reference frame, the images of the scene captured by the apparatus are filtered in a low-pass filter, so as to retain only the low spatial frequencies thereof, and the optical flow equation is solved to determine the rotations to be imposed on the images in order to stabilize them with regard to the previous images, wherein the calculation of the offsets is deduced from the steps of the stabilization process.

2. Process for correcting offsets according to claim 1, in which, after stabilization of the images, the offsets are determined via an algorithm for calculating image differences.

3. Process for correcting offsets according to claim 2, in which an offset corrections propagation algorithm is implemented.

4. Process for correcting offsets according to claim 2, in which a temporal filtering of the offsets frames is carried out .

5. Process for correcting offsets according to claim 4, in which, after filtering the offsets frames, a convergence test and a management of the length of the chains of pixels are carried out.

6. Process for correcting offsets according to claim 1, in which feedbacks of the offsets are carried out.

7. Process for detecting mobile objects in a scene by capturing snapshots with the aid of an apparatus of an imaging system whose images of the scene are stabilized according to the process of claim 1, wherein after pairwise stabilization of the images and correction of offsets by image difference calculation, the differences are summed so as to eliminate the fixed objects of the scene from the images.

8. Process for the electronic harmonization of two snapshot capturing apparatuses of two imaging systems both capturing images of the same scene, in which, in a terrestrial reference frame, the images of the scene captured at the same instants by the two apparatuses are filtered in a low-pass filter, so as to retain only the low spatial frequencies thereof, and the optical flow equation between these pairs of respective images of the two apparatuses is solved so as to determine the rotations and the variation of the relationship of the respective zoom parameters to be imposed on these images so as to harmonize them with one another.