Research For Online Print Detection System

At present, digital inkjet technology in the electronics industry is widely used, but for the e-jet printing technology, has always been inseparable from the development of the corresponding ink. For the e-jet printing ink used in its various performance parameters are higher, and it wants to develop better able to match the ink heads will not be easy. So it is necessary to timely and accurately detect the state of ink in the process of jetting.Traditional inkjet detection is using artificial observing and measurement. With the development of computer vision technology, computer detection instead of manual detection comes true. This paper makes use of moving object detection technology and digital image processing technology in printing the testing process, designs and develops a set of line print inspection system, implementing the online auto-detection with the jet parameters. The system is developed to the extent possible; to exclude artificial observation error; to quickly and accurately carry out the quantitative detection of inkjet parameters; to improve the degree of automation and intelligent on the print process.First of all, this paper introduces the design of the on-line print detection system based on the principle of inkjet technology, mainly including acquisition module, control module and detection module. Through the high speed CCD camera captures ink drop in real-time; by the control module to control the light source, the movement as well as the spray nozzles; by the detection module to achieve the core of drop detection algorithms.Then, this paper presents a detection algorithm based on the video drop. In the pre-processing stage, in order to eliminate the ignition frequency of different scene brightness of the proposed automatic brightness adjustment methods, the LED light automatically adjust the current, making the frame images collected at different frequencies to achieve basically the same brightness. Drop for high-speed movement, proposed a motion-based background subtraction drop detection method, by the background Gaussian modeling to achieve in the context of a static drop test. In order to be able to get a more accurate drop area, and further post-processing method used before the use of histogram equalization to enhance the drop area, then use region growing method to extract the frames in the exact area drop.This paper based on the extraction of drop presents a method of automatic detection of drop parameters. The speed and weight of drop are the two main parameters of jet. Through the program proposed in this paper, you can quickly and easily obtained under different conditions (such as voltage, ignition frequency, and pulse width) and weight change in the speed curve, these parameters can be directly applied to the ink nozzle and the work of state analysis. Comparing drop automatically measurement results with the traditional, the former achieves better.Finally, to achieve on-line print detection system development, given the detection of the relevant test. As an application, this paper also carried out nozzle clogging printing experiments, the method can effectively ensure the quality of the print. At present, the system has been successful in the enterprise applications, as the only domestic full-line print inspection system, its value is fully reflected.