Method for measuring the growth of leaf disks of plants and apparatus suited therefor

• US 8,391,591 B2
• Filed: 11/11/2009
• Issued: 03/05/2013
• Est. Priority Date: 12/03/2008
• Status: Active Grant

• Current Assignee: Forschungszentrum Jã¼Lich GmbH (Bundesrepublik Deutschland)
• Original Assignee: Forschungszentrum Jã¼Lich GmbH (Bundesrepublik Deutschland)
• Inventors: Biskup, Bernhard, Fischbach, Andreas, Scharr, Hanno
• Primary Examiner(s): Seth, Manav
• Application Number:
  US12/998,663
• Publication Number:
  US 20110222737A1
• Time in Patent Office: 1,210 Days
• Field of Search: None
• US Class Current: 382/154
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06T 7/62 of area, perimeter, diameter or volume

Abstract

Disclosed is a method and an apparatus for measuring the growth of leaf disks. The method comprises the following steps: a) Calibrating the capture system, b) capturing at least 2 images of a leaf disk, c) processing the image data, comprising i) segmenting the leaf disks by threshold segmentation, ii) multiple morphological erosion steps, iii) edge cleaning to remove the edge of the storage vessel, d) 3D reconstruction: generating a disparity map by means of a stereo algorithm, calculating a 3D surface model based on the previously determined calibration parameters from the disparity map, e) smoothing the previously obtained surface model, f) determining the growth rates from the time series of the areas. The apparatus comprises at least one camera, a lighting unit, a translation table in the X/Y plane for the camera and (infrared) lighting, storage vessels for leaf disks, and electronic evaluation and control units.

Claims

1. A method for measuring the growth of leaf disks, comprising the following steps:

a) calibrating a capture system and determining calibrating parameters of the capture system;

b) capturing at least one stereo image pair of a leaf disk using the calibrated capture system;

c) processing each capture image, comprising;

i) segmenting the leaf disks by threshold segmentation,

ii) carrying out morphological erosion steps on the segmented images,

iii) carrying out an edge cleaning to remove the edge of the storage vessel and detecting the edge of the leaf disk on the morphologically eroded images,

(d) determining the projected leaf area A_p of the leaf disk from the edge-cleaned images

e) carrying out a 3D reconstruction comprising;

generating a disparity image D by means of a stereo algorithm from the captured image pairs;

calculating a 3D surface model based on the previously determined calibration parameters from the disparity image;

f) smoothing the previously obtained surface model;

g) determining the 3D area A_3D of the leaf disk from the smoothed surface model;

h) repeating the image capture according to step b) at different times;

i) generating time series of the determined areas by evaluating the images that;

were captured at different times in accordance with steps c-g)

j) determining the growth rates of the leaf disk from the time series of the determined areas.

2. The method according to claim 1, wherein the calibration of the capture system at least 30 images of a checkerboard pattern are captured from different orientations, and the camera parameters of focal length, point of intersection of the optical axis with the image plane, and radial and tangential distortion are determined.

3. A method according to claim 1, wherein during the image capture two pictures are captured for each capture position, wherein the base b is selected so as to be large enough that the leaf disks are visible in the left and in the right image.

4. A method according to claim 1, wherein during the threshold segmentation, binary masks (M_t) are generated by applying a minimal and a maximal threshold value, wherein these values are determined by incrementally varying and comparing the generated binary mask to the visually recognizable boundaries of the leaf disk.

5. A method according to claim 1, wherein during the edge cleaning the increase in the pixel count or pixel area with increasing distance r from the center of gravity c_i of the leaf disk,

6. A method according to claim 1, wherein during the generation of the disparity image D by means of the stereo algorithm for each picture point in an image, the displacement of the corresponding point in the other image is determined.

7. A method according to claim 2, wherein during the image capture two pictures are captured for each capture position, wherein the base b is selected so as to be large enough that the leaf disks are visible in the left and in the right image.

8. A method according to claim 2, wherein during the threshold segmentation, binary masks (M_t) are generated by applying a minimal and a maximal threshold value, wherein these values are determined by incrementally varying and comparing the generated binary mask to the visually recognizable boundaries of the leaf disk.

9. A method according to claim 3, wherein during the threshold segmentation, binary masks (M_t) are generated by applying a minimal and a maximal threshold value, wherein these values are determined by incrementally varying and comparing the generated binary mask to the visually recognizable boundaries of the leaf disk.

10. A method according to claim 2, wherein during the edge cleaning the increase in the pixel count or pixel area with increasing distance r from the center of gravity c_i of the leaf disk,

11. A method according to claim 3, wherein during the edge cleaning the increase in the pixel count or pixel area with increasing distance r from the center of gravity c_i of the leaf disk,

12. A method according to claim 4, wherein during the edge cleaning the increase in the pixel count or pixel area with increasing distance r from the center of gravity c_i of the leaf disk,

13. A method according to claim 2, wherein during the generation of the disparity image D by means of the stereo algorithm for each picture point in an image, the displacement of the corresponding point in the other image is determined.

14. A method according to claim 3, wherein during the generation of the disparity image D by means of the stereo algorithm for each picture point in an image, the displacement of the corresponding point in the other image is determined.

15. A method according to claim 4, wherein during the generation of the disparity image D by means of the stereo algorithm for each picture point in an image, the displacement of the corresponding point in the other image is determined.

16. A method according to claim 5, wherein during the generation of the disparity image D by means of the stereo algorithm for each picture point in an image, the displacement of the corresponding point in the other image is determined.