Method and apparatus for detecting surface features of translucent objects

• US 5,845,002 A
• Filed: 11/03/1994
• Issued: 12/01/1998
• Est. Priority Date: 11/03/1994
• Status: Expired due to Term

• Current Assignee: Sunkist Growers Incorporated
• Original Assignee: Sunkist Growers Incorporated
• Inventors: Heck, Richard D., Affeldt, Henry A.
• Primary Examiner(s): Kelley, Christopher S
• Application Number:
  US08/334,165
• Time in Patent Office: 1,489 Days
• Field of Search: 382/110, 382/141, 382/142, 209/588, 209/511, 348/127, 348/128, 348/131, 348/164, 348/89, 348/91, 356/239, 356/240, 356/237, 356/52, 356/337, 356/338
• US Class Current: 382/110
• CPC Class Codes: B07C 5/3416 according to radiation transmissivity, e.g. for light, x-rays, particle radiation detecting hidden objects G01V
• CPC Class Codes: G01N 21/88 Investigating the presence of flaws or contamination

Abstract

The topographic surface features of a translucent object, such as a citrus fruit with a peel, are scanned and evaluated to permit the classification thereof according to its surface features. In the case of citrus fruit, the coarseness or pebbliness, puff and crease, ridge and valley, cuts, punctures, scrapes and splits, clear rot or sour rot of the peel is optically identified through digital analysis of the pixel scans and sorted based upon the peel surface quality. The object is classified by separating the scanned image of the fruit from the background image and removing the background image. A statistical evaluation of the image of the object as a whole, including both hemispheres of the object, is made to determine if there is any surface feature variation which might qualify as a defect or be a suitable basis upon which a classification judgment can be made. If there is, the object image is subject to high frequency or low pass filtering and thresholding of the pixel intensities to derive a refined image. The refined image is then tabulated or organized into neighborhoods contiguous to sharp transitions or gradual transitions to identify specific areas defined as blobs which, when compared against a given minimum area, shape requirement and/or width can be identified as one of the surface imperfections sought.

Claims

1. A method of classifying a translucent object having a surface with at least one of a plurality of possible surface feature patterns, said object being classified according to said at least one possible surface feature pattern, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of said at least one surface feature pattern; and

classifying the object on the basis of determination of said surface feature pattern,

wherein said step of analyzing the data comprises deriving a statistical measure of the data characteristic of said at least one surface feature pattern in order to identify said at least one surface feature pattern from among said plurality of surface feature patterns possible for said object,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said object is illuminated with a source having an optical beam and said step of optically forming an image comprises optically reflecting an image of only the scattered light transmitted from the object into a camera pointed away from the object and not in said optical beam of said source.

2. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein a plurality of surface features are considered, and said step of analyzing comprises determining if any of such surface features exists on the object,

wherein the object is a citrus fruit with a peel, and wherein said step of analyzing comprises determining the extent of porosity of the peel of the citrus fruit.

3. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein a plurality of surface features are considered, and said step of analyzing comprises determining if any of such surface features exists on the object,

wherein the object is a citrus fruit with a peel, and wherein said step of analyzing comprises determining the extent of puff and crease in the peel of the citrus fruit.

4. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein a plurality of surface features are considered, and said step of analyzing comprises determining if any of such surface features exists on the object,

wherein the object is a citrus fruit with a peel, and wherein said step of analyzing comprises determining the extent of ridges and valleys in the peel of the citrus fruit.

5. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein a plurality of surface features are considered, and said step of analyzing comprises determining if any of such surface features exists on the object,

wherein the object is a citrus fruit with a peel, and wherein said step of analyzing comprises determining the extent of fractures in the peel of the citrus fruit.

6. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein a plurality of surface features are considered, and said step of analyzing comprises determining if any of such surface features exists on the object,

wherein the object is a citrus fruit with a peel, and wherein said step of analyzing comprises determining the extent of surface decomposition in the peel of the citrus fruit.

7. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said step of analyzing comprises;

separating the image into a background image and an object image; and

removing the background image, leaving an object image,

wherein the object is a fruit with a peel and wherein said step of analyzing comprises determining the extent of porosity of said peel, and further comprising;

scanning the object image;

filtering the scanned image for high frequency data using a high pass data filter;

establishing a threshold of pixel intensity of the filtered image; and

tabulating a number of pixels having an intensity above the threshold.

8. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said step of analyzing comprises;

separating the image into a background image and an object image; and

removing the background image, leaving an object image,

wherein the object is a fruit with a peel and wherein said step of analyzing comprises determining the existence of hills and valleys in the peel, and further comprising;

scanning the object image;

filtering the scanned image for low frequency data using a low pass/smoothing data filter;

establishing a threshold of pixel intensity of the filtered image; and

tabulating the number of pixels having an intensity above the threshold.

9. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said step of analyzing comprises;

separating the image into a background image and an object image; and

removing the background image, leaving an object image,

wherein the object is a fruit with a peel and wherein said step of analyzing comprises determining the existence of raised and recessed wedges in the peel, and further comprising;

selectively scanning the object image for sharp transitions in pixel intensity;

establishing a threshold of pixel intensity;

tabulating the number of pixels having an intensity above the threshold in neighborhoods contiguous to pixel boundaries corresponding to sharp transitions, said neighborhoods being defined as blobs; and

tabulating the number of blobs which exceed a predetermined characteristic parameter, whereby raised and recessed wedges in the peel are identified.

10. The method of claim 9 wherein the characteristic parameter comprises a selected minimum blob area, and said step of tabulating the number of blobs comprises tabulating the number of blobs exceeding the selected minimum area.

11. The method of claim 9 wherein the characteristic parameter comprises a selected blob shape, and said step of tabulating the number of blobs comprises tabulating the number of blobs complying with the selected shape definition.

12. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said step of analyzing comprises;

separating the image into a background image and an object image; and

removing the background image, leaving an object image,

wherein the object is a fruit with a peel and wherein said step of analyzing comprises determining the existence of fractures in the peel surface, and further comprising;

scanning the object image for sharp transitions and pixel intensity;

establishing a selected threshold of pixel intensity of the object image; and

tabulating the number of pixels having an intensity above the selected threshold in neighborhoods continuous to pixel boundaries with sharp transitions, such neighborhoods being defined as blobs; and

tabulating the number of such blobs which exceed a selected area and width, whereby fractures in the peel are identified.

13. A method of classifying a translucent object according to at least one selected surface feature of the object, comprising the steps of:

illuminating the object to cause light to be scattered therewithin;

detecting such scattered light transmitted through the surface of the object;

converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface;

analyzing such data to determine the existence of the selected feature of the surface; and

classifying the object on the basis of such determination,

wherein the step of detecting further comprising the step of optically forming an image of the transmitted scattered light from the object, and

wherein said step of analyzing comprises;

separating the image into a background image and an object image; and

removing the background image, leaving an object image,

wherein the object is a fruit with a peel and wherein said step of analyzing comprises determining the existence of surface decomposition, and further comprising;

scanning the object image for gradual transitions in pixel intensity;

establishing a selected threshold of pixel intensity;

tabulating the number of pixels having an intensity above the selected threshold in neighborhoods contiguous to pixel boundaries with gradual transitions, such neighborhoods being defined as blobs; and

tabulating the number of such which exceed a given characteristic parameter,

whereby surface decomposition is identified.

14. The method of claim 13 wherein the characteristic parameter comprises a selected minimum area of such blob, and said step of tabulating comprises tabulating the number of blobs which exceed the selected minimum area.

15. The method of claim 13 wherein the characteristic parameter comprises a selected shape, and said step of tabulating comprises tabulating the number of blobs having a shape complying with the selected shape definition.

16. An apparatus for identifying at least one selected surface feature pattern on a translucent object, said object having a plurality of possible surface feature patterns, comprising:

means for illuminating the object to cause light to be scattered therewithin;

means or detecting such scattered light transmitted through the surface of the object;

means for converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface; and

means for analyzing said data to derive a statistical measure of said data to identify said one of said surface feature patterns on said surface of said object,

wherein said means for analyzing said data to derive a statistical measure of said data comprises means for forming a stored two dimensional digital image of said object, means for scanning said stored two dimensional digital image along a at least one straight line to obtain a scan of image intensity, means for generating a histogram of image intensity of said stored two dimensional digital image along said at least one straight line, and means for comparing said histogram to predetermined criteria to identify said one of said surface feature patterns.

17. The apparatus of claim 16 wherein said means for scanning scans said stored two dimensional digital image along a plurality of straight lines and averages said corresponding plurality of scans to obtain a scan of average image intensity from which said histogram is generated.

18. An apparatus for identifying at least one selected surface feature pattern on a translucent object, said object having a plurality of possible surface feature patterns, comprising:

means for illuminating the object to cause light to be scattered therewithin;

means for detecting such scattered light transmitted through the surface of the object;

means for converting such detected light to data representing the relative intensity of light transmitted through each of a plurality of defined areas of the surface; and

means for analyzing said data to derive a statistical measure of said data to identify said one of said surface feature patterns on said surface of said object,

wherein said means for detecting scattered light comprises a scanning camera directed at said object, but not in a line of sight with said means for illuminating said object, said scanning camera being otherwise shielded from said means for illuminating said object so that said scanning camera receives only signals scattered through said object and not from direct or reflected rays from said means for illuminating said object.

19. The apparatus of claim 18 wherein said scanning camera is turned diametrically away from said means for illuminating said object and further comprising an assembly of mirrors for reflecting only signals scattered through said object to said scanning camera.