Torna alla pagina italiana | Back to the English page

ODILE

On Line Defects Identification on the Loom Equipment

List of ODILE Reports by CEO and INO.

The main purpose of the ODILE project was to develop a device performing

WEAVING DEFECT DETECTION BY FOURIER IMAGING

Fig. 2	Fig. 3

In the framework of the ODILE project we have studied and tested

DEFECT DETECTION IN TEXTURED MATERIALS BY OPTICAL FILTERING WITH STRUCTURED DETECTORS AND SELF ADAPTABLE MASKS

Fig. 4

An optical method using Fourier transformation and spatial filtering is used to reveal defects in textured materials in real time. New optical structured filter types were developed including a self adaptable mask made of photocromic polymers. The characteristics of these materials allow very promising applications in pattern recognition such as represented by a fabric. Fig. 4 presents the experimental system to evaluate variations in the O.F.T. (Optical Fourier Transform). Fig. 5 shows the O.F.T. image of the fabric on a filter with concentring rings. While the O.F.T. image on a "double butterfly" filter is presented in Fig. 6.

Among the possible materials that can be used to realise the masks, some photochromic materials appear to be very promising. Typically photochromic molecules are unstable: the activated state lasts few minutes and presents inconstant characteristics. There is only a class of photochromic molecules, called Fulgides, which presents two stable states. Fulgides are activated by UV light and their colour becomes darker, then they can be disactivated by visible light.

For the realisation of ODILE masks the Fulgides needed to be incorporated in a solid matrix, which should not corrupt their photochromic characteristics.

CHARACTERISATION OF FULGIDES IN A SOLID MATRIX AS RECORDING MATERIALS FOR OPTICAL COMPONENTS

Fig. 7

The optical properties of solid samples of polymethyl-methacrylate doped with fulgide photochromic molecules were studied. The transmittance of a thin plate sample was measured when both the activation (UV) and bleaching wavelengths are employed. The samples were used as an optical switch where an UV pulse switches a visible signal beam. The behaviour of these compounds as holographic materials, studied by efficiency measurements at varying grating steps, is reported.

Fig. 7 shows the experimental set-up for the measurements in competition with simultaneous UV and visible irradiation on the photochromic sample. Fig. 8 presents the transmittance of the fulgide 611TD in polymethil-methacrylate measured for increasing value of the input visible light, the five curves refer to different levels of the UV irradiation on the sample.

Finally a different approach for the identification of defects in OFT image has been investigated:

NEURAL NETWORKS IN THE OPTICAL RECOGNITION OF DEFECTS IN CLOTH

A fast system to reveal the presence and type of fabric defects during the weaving process has been developed. Since the fabric is similar to a two-dimensional grid, its defects are clearly observed in the changes in its Optical Fourier Transform (OFT), which appears stationary while the fabric is moving across the loom. Previous works, based on the statistical parameters of the OFT, show that the presence of faults can be detected when only global changes in the images are considered. This paper shows that by selecting a small subset of pixels from the image as input to a neural network, fabric defects can not only be detected but also successfully identified. A knowledge-based system could conceivably be constructed to use this information to resolve problems with the loom in real-time, without the need for operator intervention.

Fig. 9 presents the OTF of a sample of defected-free cotton cloth. The subset in the white box was used as imput in the neural network. Performances of the network are show in Fig. 10.

Torna alla pagina italiana | Back to the English page