Multi-spectral spectroscopic photography-based visual monitoring system

Abstract

The invention relates to a multi-spectral spectroscopic photography-based visual monitoring system. The multi-spectral spectroscopic photography-based visual monitoring system is characterized by comprising a camera, wherein more than one auxiliary light sources are arranged outside the camera; a semi-transmissive and semi-reflective lens, a first narrow band filter, a first dimming element, a first area array image sensor, a second narrow band filter, a second dimming element and a second area array image sensor are arranged in the camera; light emitted from the auxiliary light source irradiates a welding workpiece; the light rays reflected by the welding workpiece are transmitted to the semi-transmissive and semi-reflective lens through the lens of the camera; the light rays transmitted from the semi-transmissive and semi-reflective lens are transmitted to the first area array image sensor through the first narrow band filter and the first dimming element in turn to complete imaging; and the light rays transmitted from the semi-transmissive and semi-reflective lens are transmitted to the second area array image sensor through the second narrow band filter and the second dimming element in turn to complete imaging. The multi-spectral spectroscopic photography-based visual monitoring system can be widely applied to automatic welding bead recognition processes of groove detection, welding seam tracking, non-destructive testing after welding and the like of a welding robot or other automation equipment.

Description

technical field [0001] The present invention relates to an arc welding process visualization system, in particular to a visual monitoring system based on multi-spectral spectroscopic photography for simultaneous observation of multiple objects such as front grooves, molten pools, welding wires, and arcs. Background technique [0002] The visual monitoring of the arc welding process refers to the observation of the welding area during the real-time welding process. This process is of great significance for controlling and stabilizing the welding quality. The observation objects mainly include: front groove, molten pool, welding wire, welding arc, etc. In the prior art, the use of cameras to shoot the arc welding process for visual monitoring plays an important role in production: 1. When performing welding operations in places that are difficult or unsuitable for people to reach (such as: high altitude, underwater, dangerous environments, etc.), the welding Remote monitoring ...

AI Technical Summary

Problems solved by technology

Considering the above-mentioned characteristics of arc luminescence, the solution to the visualization problem of the near-arc area in the arc welding process monitoring in the prior art is as follows: (1) Install a narrow-band filter and a light reduction device in front of the camera to filter out the light to the greatest extent. Under the influence of strong arc light, keep the spectral segment of interest (for example: when using laser structured light with a wavelength of 640nm or 660nm for visual tracking of weld seams, select a narrow-band filter with the corresponding central wavelength to allow only light near the central wavelength to pass through, to achieve The purpose of reducing arc interference), but because only a single spectrum image can be obtained, the information is not rich enough to complete the acquisition of multi-feature information; (2) A dual-camera system is used, and one of the cameras shoots the molten pool area through a filter , the light intensity of the molten pool is weakened, and the profile information of the molten pool can be obtained after processing. Another camera directly shoots the front groove without a filter, which is used to observe the groove position under natural light conditions before welding starts. , the disadvantage of this method is that: the two cameras have different fields of view, and there is an image registration problem between the obtained images, which is easy to introduce errors, and it is impossible to obtain a clear and stable front groove image during the welding process; (3) use Pulsed laser or high-brightness flashlight is used as the lighting source to intermittently illuminate the welding area with high brightness. At the same time, the high-speed shutter of the camera is controlled to collect images at the moment when the welding area is illuminated by the light source. Although this method can make the molten pool in the welding area The brightness of the surrounding workpiece can be improved, but also because of the use of a single camera and the only optical path, only a single spectrum image can be captured, and because the lighting source with a brightness higher than that of the welding arc must be used, it increases the difficulty of equipment manufacturing and installation volume, which is not conducive to control cost

[0004] To sum up, the various parts that need to be observed during the arc welding process correspond to different spectral regions, and the light entering the camera in these different spectral regions has a huge difference in light intensity. The method of spectral image acquisition, or the method of using dual cameras to filter light in different bands (or adding a filter to one camera and not using a filter to the other camera), cannot achieve different spectral regions from the same field of view. The light adopts different filter and dimming measures

Images