TWI553186B - Classification method of fabric categories - Google Patents

Description

Method for grading cloth grade

The present invention relates to a method of classifying grades, and more particularly to a method of classifying cloth grades.

Taiwan's textile industry was mainly based on imported raw materials, and was transferred to the petrochemical industry to extract raw materials. Recently, it has actively developed man-made fibers, and combined with the concept of reinsurance, using coffee slag or pudding bottles for recycling and chopping. The material of man-made fiber not only reuses waste, but also reduces environmental pollution. However, such product innovation and innovative equipment investment create high value-added products, but traditional artificial visual methods, artificial The visual method can only clearly see the hair ball, it is difficult to see the pile group (not composed of the hair ball), because the pile group is not solid and opaque like the whole hair ball, so the manual method is difficult to accurately classify the cloth. Accurate observation of the lack of fluffy groups, not only the level of classification has not improved, it is easier to have a wrong classification.

Furthermore, even if some parts are discriminated by photography, direct illumination is generally used in the light source part. However, since the light source directly illuminates (excessively) and easily penetrates the pile group, the light source cannot be stagnated on the surface of the pile, and the camera cannot. It is also necessary to improve the filming of the villi.

Therefore, if a set of objective and uniform standards can be developed The level method, without losing the surface roughness of the cloth as the basis for judging, will greatly improve the precision of the cloth grade of the textile industry and improve the classification efficiency.

Accordingly, it is an object of the present invention to provide a method for grading cloth grades, thereby enabling rapid and accurate grading of cloth to achieve high grading quality and grading efficiency.

Therefore, the method for grading the cloth grade of the present invention includes a pre-step, an image capturing step, an image processing step, an operation step, a grading step, and the like; wherein, in the pre-step, the cloth to be tested is partially taken and transmitted through After the surface of the cloth is rubbed against each other, the image capturing step obtains an image for the rubbing portion of the cloth, and the image processing step processes the image to obtain a black dot, so that the operation step will follow the black dot. Calculating the number of a ball, and comparing the number of balls to the area of the piece to be taken, and then calculating a ball value. Finally, the step of grading is the value of the ball in the operation step and the step of the step. The value of the hairballs in the database is compared so that the grade of the hairball can be graded according to the value of the hairball; thus, the detection and classification of the human visual inspection can be replaced, and the lack of error due to visual observation can be avoided, thereby achieving rapid And accurate classification quality and efficiency; of course, the light source in the image capturing step may be an indirect light source to indirectly illuminate the cloth to be tested Surfaces, and thus clearly illustrating that the pile group respectively (not constituting the hair bulb), to achieve a more accurate grading effect.

3‧‧‧ Method of grading cloth grade

31‧‧‧Pre-steps

32‧‧‧Image capture steps

33‧‧‧Image processing steps

34‧‧‧Operation steps

35‧‧‧Classification steps

331‧‧‧ Grayscale program

332‧‧‧ threshold adjustment procedure

333‧‧‧ pixel selection procedure

35‧‧‧Classification steps

351‧‧‧Database

352‧‧‧Processing unit

1 is a block diagram of a preferred embodiment of the present invention.

2 is a flow chart of a preferred embodiment of the present invention.

FIG. 3 is a schematic diagram showing the result of the grayscale process in the image processing step of the present invention.

4 is a schematic diagram showing the result of the threshold adjustment procedure in the image processing step of the present invention.

FIG. 5 is a schematic diagram showing the result of a pixel selection procedure in the image processing step of the present invention; FIG.

The above and other technical features, features, and advantages of the present invention will become apparent from the Detailed Description of the <RTIgt;

Referring to FIG. 1 and FIG. 2, in a first embodiment of the present invention, a method 3 for grading a cloth level in the embodiment includes a pre-step 31, an image capturing step 32, an image processing step 33, and an operation. Step 34 and grading step 35; wherein the pre-step 31 is provided with a cloth to be tested, and a part of the block to be tested is taken, and the surface of the cloth is subjected to rubbing treatment, such as through a wear meter. The image capturing step 32 is performed by using a camera (not shown) and assisted by at least one light source (not shown), such as the camera in the embodiment. In the CCD camera lens, the aforementioned cloth to be tested is subjected to image capturing of a portion of the cloth area through the friction portion, thereby obtaining an image. Furthermore, in the embodiment, the light source is mainly an indirect light source for indirectly illuminating the image. The surface of the cloth to be tested, so that the fluffy group (not composed of the hair ball) on the rough surface of the cloth can be clearly displayed, and the number of hair balls can be revealed.

Continuing the foregoing, the image processing step 33 includes a grayscale program 331, a threshold adjustment program 332, and a pixel selection program 333, wherein The grayscale processing program 331, in this embodiment, performs the grayscale (8-bit) processing on the image obtained as described above, and then the black and white colors are reversed for observation, as shown by the black dots in FIG. For the pile or the hair ball, the threshold adjustment program 332 will adjust the threshold to eliminate unnecessary tiny black spots to form the state of FIG. 4, and the pixel selection program 333 sets the black point. The minimum pixel is set to be less than the number of pixels below the black point to be removed, and the black point below which the number of pixels is left, which has been formed as shown in FIG.

Furthermore, the operation step 34 calculates a number of balls according to the black point, and compares the number of the balls with the area of the captured pieces, and in the embodiment, the number of balls used/taken The cloth area is further calculated as a gross value; finally, the grading step 35 includes a database 351 having a hairball value of each level, and the operation step 34 and the hair ball in the database 351, respectively. The value is compared with the processing unit 352, wherein in the embodiment, the gross value in the database 351 can be set to a grace range, for example, the value of the hairball value is within the range, and then the level is set, and then When the processing unit 352 reads the gross value in the operation step 34, it can be compared with the gross value in the database 351, and then the level of the cloth to be tested is graded, as shown in the figure. And dividing into one to five levels; in order to sequentially pass the pre-step 31, the image capturing step 32, the image processing step 33, the calculating step 34, and the stepping step 35, the cloth can be graded one by one. Grading to replace traditional visual inspection Method effectively achieve accurate classification of fabric, the fabric and the speed of the classification efficiency is more improved.

In the foregoing, the method for grading a cloth grade of the present invention comprises a pre-step, an image capturing step, an image processing step, an operation step, and a grading step, wherein a part of the cloth to be tested is taken and rubbed against the surface of the cloth. (ie, the pre-step), and after obtaining an image for the rubbing portion of the cloth (ie, the image capturing step), the image is processed to obtain the most appropriate black dot (ie, the image processing step), Then, the number of balls is calculated according to the black point, and the number of the balls is compared with the area of the captured pieces to calculate a ball value (that is, the operation step), and finally, the classifying step is The value of the hairball in the calculation step is compared with the value of the hairball in the database in the classification step, so that the grade of the hairball is graded according to the value of the hairball, thereby achieving fast and accurate classification quality and efficiency, so it is true The object of the invention is achieved.

The above is only the preferred embodiment of the present invention, and the scope of the present invention is not limited thereto, and the simple equivalent changes and modifications made in accordance with the scope of the present invention and the contents of the invention are generally modified. All should remain within the scope of the invention patent.

3‧‧‧ Method of grading cloth grade

31‧‧‧Pre-steps

32‧‧‧Image capture steps

33‧‧‧Image processing steps

34‧‧‧Operation steps

35‧‧‧Classification steps

331‧‧‧ Grayscale program

332‧‧‧ threshold adjustment procedure

333‧‧‧ pixel selection procedure

35‧‧‧Classification steps

351‧‧‧Database

352‧‧‧Processing unit

Claims (2)

A method for grading a cloth grade, comprising: a pre-step, which has a cloth to be tested, and takes a part of the cloth to be tested, and rubs the surface of the cloth to be tested with each other; a step of using a camera and assisting at least one light source, wherein the cloth to be tested passes through the rubbing portion to perform image capturing of a portion of the cloth area, thereby obtaining an image; and an image processing step, which includes a grayscale in sequence a program, a threshold adjustment program and a pixel selection program, wherein the grayscale program performs grayscale processing on the image obtained as described above, so that the hairball forms a majority of black spots, the threshold adjustment program, and the gray The stepped image is adjusted to be a small black dot, and the pixel selection program sets the black dot to be left with the allowable black dot pixel size; an operation step, which calculates a hairball according to the black dot a quantity, and comparing the number of the balls to the area of the piece to be taken, thereby calculating a ball value; and a grading step including including one of the balls of each level a library, and a processing unit that respectively compares the operation step with a hairball value in the database, through which the hairball value in the operation step is compared with the hairball value in the database, In order to rank the grade according to the value of the hairball. The method according to claim 1, wherein the light source is an indirect light source to indirectly illuminate the surface of the cloth to be tested.