织物图像分析软件设计及与超宽幅扫描仪的配套应用

发布时间：2018-03-10 09:03

本文选题：数字图像　切入点：TWAIN　出处：《浙江理工大学》2013年硕士论文　论文类型：学位论文

【摘要】：扫描仪是一种把模拟影像转化为数字影像的光机电一体化设备产品，广泛应用于出版、印刷、广告制作、图文通讯等多个领域。而在纺织工业领域中，纺织品幅面一般较大，普通宽幅扫描仪及其配套软件大多无法满足织造行业的尺寸要求，因此，必须使用超宽幅面的扫描仪对织物进行图像扫描。当前，国内市场上超宽幅面的扫描仪数量相对较少，配套软件资源也相对匮乏，在此市场背景下，设计研发了一款超宽幅扫描仪良好的解决了纺织品的扫描问题，为此超宽幅扫描仪开发一款与之配套的软件系统正是本课题的研究目的所在。软件系统通过编写遵循TWAIN协议的标准接口程序将织物静态图像从超宽幅扫描仪上传输到应用程序中，实现计算机与超宽幅扫描仪的图像数据传输。然后通过加权平均、中值滤波等方法对获取的织物图像进行灰度化、消除噪声等预处理，并在此基础上利用空间坐标变换、像素灰度内插等方法对织物图像获取过程中出现的失真图像加以几何校正。最后，综合利用Hough变换、二值化处理、Sobel算子细化等一系列图像处理和识别方法对织物图像进行边缘检测，实现对纺织产品质量的控制，增强了整个系统的实用性。此外，为了提高软件系统的适用性，在配套软件中添加了图像局部细节放大、图像旋转、格式转换等图像查看功能以及图像自由裁剪、画笔、橡皮擦等图像编辑工具。在初步试验中，应用程序运行效果良好，，在TWAIN协议的保障下能够快速地从超宽幅扫描仪上获取织物图像，并可以准确显示织物疵点信息，结合其他功能模块设计，应用程序提供了一个方便的可视化操作界面，使技术人员可以及时了解产品质量问题，并可通过程序相关设计对织物图像进行编辑。扫描检测技术应用前景广阔，但是，目前国内超宽幅扫描仪市场尚未成熟，织物检测关键技术的研究状况仍处于初级阶段，尚需进一步提高自动检测技术水平。
[Abstract]:Scanner is a kind of optical-electromechanical equipment which converts analog image into digital image. It is widely used in publishing, printing, advertisement making, graphic and text communication and other fields. Most of the general wide-width scanners and their supporting software can not meet the size requirements of the weaving industry, so we must use ultra-wide-format scanners to scan the fabrics. At present, the number of ultra-wide-format scanners in the domestic market is relatively small. The matching software resources are also relatively scarce. Under this market background, the design and development of an ultra-wide scanner has solved the problem of textile scanning well. Therefore, the purpose of this research is to develop a matching software system for ultra-wide scanners. The software system transmits the fabric static image from the ultra-wide scanner to the application program by writing the standard interface program following the TWAIN protocol, and realizes the image data transmission between the computer and the ultra-wide scanner. On the basis of the method of median filter, the fabric image is grayscale, noise is eliminated, and the space coordinate transformation is used. Pixel grayscale interpolation and other methods are used to geometric correct the distorted image in the process of obtaining fabric image. Finally, the Hough transform is used synthetically. In order to improve the applicability of the software system, a series of image processing and recognition methods such as binary processing and Sobel operator thinning are used to detect the edges of fabric images, which can control the quality of textile products and enhance the practicability of the whole system. Image viewing functions such as image local detail amplification image rotation format conversion and image editing tools such as image cutting brush eraser and other image editing tools are added in the matching software. In the preliminary test, the application program works well. Under the guarantee of TWAIN protocol, the fabric image can be quickly obtained from the ultra-wide scanner, and the fabric defect information can be accurately displayed. The application can be designed with other functional modules. The application program provides a convenient visual operating interface, which enables technicians to understand product quality problems in a timely manner, and can edit fabric images through program related design. At present, the domestic market of ultra-wide scanners is not mature, the research status of the key technology of fabric detection is still in the primary stage, and it is necessary to further improve the level of automatic detection technology.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP334.22;TP391.41

【参考文献】