Process for detecting surface defects on a textured surface

• US 6,040,853 A
• Filed: 10/02/1997
• Issued: 03/21/2000
• Est. Priority Date: 10/02/1996
• Status: Expired due to Fees

• Current Assignee: Laboratoire Central Des Ponts et Chaussees, L'Institut de Recherche et D'Enseignement Superieur AUX Techniques de Electronique
• Original Assignee: Laboratoire Central Des Ponts et Chaussees, L'Institut de Recherche et D'Enseignement Superieur AUX Techniques de Electronique
• Inventors: Delagnes, Philippe, Barba, Dominique
• Primary Examiner(s): Britton, Howard
• Assistant Examiner(s): VO, TUNG T
• Application Number:
  US08/957,987
• Time in Patent Office: 901 Days
• Field of Search: 348/125, 348/127, 348/128, 348/132, 348/134, 348/88, 348/148, 348/150, 348/92, 382/149, 382/1, 382/152, 382/205, 382/141, 382/110, 382/8, 356/445, 356/237, 73/104, 73/105, 73/146
• US Class Current: 348/128
• CPC Class Codes: G01C 7/04 involving a vehicle which moves along the profile to be traced
• CPC Class Codes: G01N 21/8903 using a multiple detector array

Abstract

A process for detecting defects in a textured suface, particularly on a highway. The process consists in displacing with respect to the surface a camera presenting an inclination and in processing by processing systems the successive images taken by the camera. A mathematical processing is effected on sub-images, corresponding to the same spatial portion of the surface, taken at different consecutive instants.

Claims

1. Process for detecting surface defects in a substantially plane textured surface,

wherein it comprises the following steps of:

a) making, with the aid of an optical sensor presenting an inclination of sight with respect to the substantially plane textured surface and being in relative movement with respect to this surface, a succession of images of parts of this surface at successive instants, each image corresponding to a surface part of length L in the direction of displacement, each image being offset from the preceding one by a length 1, 1 being equal to L/K with K, a whole, greater than 2;

b) defining sub-images corresponding to elementary portions of the surface whose length is equal to 1 in the direction of displacement, whereby for the length of portion of surface 1, K sub-images of this portion are available;

c) extracting from these K consecutive sub-images N sub-images (Ii) corresponding to the same portion of the surface of length 1 at N successive instants with N included between 2 and K;

d) applying to each of the N sub-images (Ii) the same spatial mathematical processing in order possibly to demonstrate a configuration of presumed defect, which gives a processed intermediate sub-image (Yi);

e) applying to the first processed elementary sub-image (Yo) a combinatorial algorithm for detection of presumed defect in the elementary sub-image, whereby a spatial and temporal sub-image ST₀ is obtained;

f) iteratively repeating the application of said combinatorial algorithm to the spatial and temporal sub-image ST_i-1 obtained at instant ti-1 and to the processed elementary sub-image Yi in order to obtain a spatial and temporal sub-image ST_i up to i=N-1; and

g) obtaining with the spatial and temporal sub-image ST_N-1 the image of the elementary portion of the textured surface with its possible defects.

2. The process of claim 1, wherein, before step a), there is applied to each sub-image a processing for at least partial correction of the effects of perspective due to the inclination of the optical sensor with respect to the textured surface.

3. The process of claim 2, wherein said spatial mathematical process of step d) consists in applying to said elementary image a markovian model with regular or irregular grid.

4. The process of claim 2, wherein said spatial mathematical processing of step d) consists in effecting an extraction of the related components from the pixels of the sub-image.

5. The process of any one of claims 1 to 4, wherein said optical sensor is a camera mounted on a vehicle and the textured surface is the surface of a highway.

6. The process of any one of claims 1 to 4, wherein said optical sensor is fixed and the textured surface moves linearly with respect to the optical sensor.

7. The process of claim 6, wherein the surface is a web of fabric.

8. The process of claim 6, wherein said surface is a piece of rough wood.

9. The process of claim 6, wherein said surface is that of a long metal product.