Measurement Based On Image Processing For Detecting Water's Diffusion Capability In Fabric

The measurement based on image processing is investigated to detect the behavior of water in textiles. In the mean time the rules of them are discussed.By many image processing ways it settles the ill influence of the texture, pictures on fabric and the light, wipes off the noise in images, solves the problems which are arose by light's unbalance and dark image, improves the quality of those images. It uses a model of distortional image to rectify the images, improves the images' condition and the test's precision. A simple way is also found to calculate the equivalent focus. The precise means that uses two dots to calculate the distance of two pixels is selected by comparing two methods. They improve the precision of the apparatus.25 samples were selected to do some tests and research in the rule of the liquid's behavior in fabric. It also analyses the error and prove that the method is available. On the base of the test datum, it does many regression analysis to investigate the factors impacting the behavior of liquid in textiles and the relationship between strip test and spot test. The mathematic model of that is discussed. Through the researches and analysis, the following results can be derived from the experiments:1. The behavior of liquid in fabrics can be memorized by computer vision, and the speed can reaches 0.1 second per image. Image processing is used to get liquid's spreading area in textiles. The precision can reach 0.01 mm~2. They enhance the traditional measurement.2. The techniques that are used settle two difficult problems from the unequal brightness of images and the pattern on fabric. The special program's arithmetic can settles the bad influence from the unstable light and resolves the problem that hardwares can settle hardly3. By building relationship between pixel and area with two dots and the means of getting the lens' equivalent make operation easy.4. The arithmetic is good to rectify the distortion image and enhances precision. The sheet for pixel coordinates of the distortional image and the rectified image reduce the time of image processing.5. The curve of the liquid diffusing in textile is first got. The curve accords with the existing rule, settles the different behavior of liquid in warp and in weft and is related with strip test.6. The parameters are irrespective. They can describe the different aspect of water in textile. Selecting them to show water's behavior in fabric is right.7. The behavior of water in textile has good relations with the fabric's thickness, warp density and weft density. The behavior accords with the existing rules.8. The velocity of water in fabric has good correlation with the fabric's density. The density is bigger, the velocity is lower 9. The height and speed of water in plumb fabric have relations with the density in 10cm. The density is bigger in the same fabric, the height and speed of water in it are bigger too.10. The speed of a drip in fabric has good relations with the speed of water in plumb fabric, The product of the speeds of water along warp and weft in the same textile is bigger, the velocity of a drip in the textile is bigger too.