- about us
- contact us
in the field of industrial production, robot inspection products rely on machine vision to a large extent. the sensitivity of vision will directly affect the inspection speed and inspection quality of the product. therefore, it is particularly important to design a high-quality vision product. in the design process, the designer will facing many problems such as visual positioning, measurement, detection and recognition.
1. stability of lighting
industrial vision applications are generally divided into four categories: positioning, measurement, detection and recognition. among them, measurement requires high stability of light, because as long as the light changes by 10-20%, the measurement result may deviate by 1-2 pixels , this is not a software problem. this is a change in illumination, which has caused the edge position on the image to change. even the most powerful software cannot solve the problem. it is necessary to eliminate the interference of ambient light from the perspective of system design, and at the same time ensure active lighting luminous stability of the light source. of course, increasing the resolution of the hardware camera is also a way to improve accuracy and resist environmental interference. for example, the spatial size of the previous camera counterpart was 1 pixel 10um, but by increasing the resolution, it becomes 1 pixel 5um. the accuracy can be considered to be approximately doubled, and the interference to the environment is naturally enhanced.
two, the inconsistency of the position of the workpiece
generally speaking, whether it is offline testing or online testing, as long as it is a fully automated testing equipment, the first step to do is to find the target to be tested. every time the object to be measured appears in the shooting field of view, you must be able to know exactly where the object to be measured is. even if you use some mechanical fixtures, etc., you cannot ensure that the object to be measured appears in the same position every time. yes, this requires the positioning function. if the positioning is not accurate, the position of the measurement tool may be inaccurate, and the measurement results sometimes have large deviations.
generally, the following calibrations are required for high-precision measurement: front, optical distortion calibration (if you are not using a software lens, you must generally calibrate); second, projection distortion calibration, that is, because of your installation position error representative correction of image distortion, calibration of three-object image space, that is, to calculate the size of the corresponding object space for each pixel.
however, the current calibration algorithms are all based on plane calibration. if the physics to be measured is not plane, the calibration will need to be processed by some special algorithms, which cannot be solved by normal calibration algorithms.
in addition, for some calibrations, a special calibration method must be designed because the calibration board is not used. therefore, the calibration may not be completely solved by the existing calibration algorithm in the software.
fourth, the speed of the object
if the object to be measured is not stationary, but in motion, then we must consider the effect of motion blur on the image accuracy (fuzzy pixels = object motion speed * camera exposure time), which cannot be solved by software.
five, the measurement accuracy of the software
in the measurement application, the accuracy of the software can only be considered in terms of 1/2-1/4 pixels, preferably 1/2, and cannot achieve 1/10-1/30 pixel accuracy like positioning applications, because in measurement applications the software can extract very few feature points from the image.
the movement speed and measurement accuracy of machine vision occupies an important position in the entire product. the speed of movement is inversely proportional to the detection ability. the faster the movement, the quality of the detection is relatively poor. therefore, it is important to improve the movement accuracy and detection details.