視覺自動化檢測原理：精確檢測產品裂紋與劃痕
發布時間：
2023-02-10 11:26
在機器視覺自動檢測技術中最大的技術調整就是檢測物品表面的裂紋、劃痕、變色、燒蝕等表面缺陷。難點在于形狀不規則，深度與光線對比度低，經常受到產品表面自然紋理或圖案的干擾。
In the machine vision automatic inspection technology, the biggest technical adjustment is to detect the surface defects such as cracks, scratches, discoloration, ablation and so on. The difficulty lies in the irregular shape, low depth and light contrast, which is often disturbed by the natural texture or pattern of the product surface.
因此，表面缺陷的檢測需要高精度的照明、攝像機分辨率、被檢測部件與攝像機的相對位置以及復雜的機器視覺自動檢測技術。
Therefore, the detection of surface defects requires high-precision lighting, camera resolution, the relative position between the detected parts and the camera, as well as complex machine vision automatic detection technology.
光照是產品表面缺陷檢測中一個極其重要的因素，因為表面缺陷通常具有低對比度的特點，使用正確的光照可以將其與背景分離。對于平面凸面缺陷，可以使用低角度照明來“照亮”凸面部分的邊緣。
Illumination is an extremely important factor in the detection of surface defects of products, because surface defects usually have the characteristics of low contrast and can be separated from the background by using the correct illumination. For planar convex defects, you can use low angle lighting to "illuminate" the edges of convex parts.
水平檢測功能（玻璃盤）
水平螺牙檢測功能
頭部檢測功能（分度盤）
漫射光可用于檢測光滑缺陷或曲面缺陷。在采集的圖像中，裂紋或劃痕應至少覆蓋三個或更多像素。在某些情況下，多個高分辨率相機或單個相機用于多個拍攝，或者更常見的是，使用1K到16K的線掃描相機以指定的分辨率檢測零件的整個表面，以滿足用戶的精度要求。
Diffuse light can be used to detect smooth defects or surface defects. Cracks or scratches should cover at least three or more pixels in the captured image. In some cases, multiple high-resolution cameras or a single camera are used for multiple shots, or more commonly, a line scan camera from 1K to 16K is used to detect the entire surface of the part at a specified resolution to meet the user's accuracy requirements.
為了利用機器視覺自動檢測技術檢測產品的裂紋和劃痕，首先要了解產品表面缺陷是如何產生的，并在圖像中清晰地分類這些缺陷可能的形狀。例如，粉末金屬零件材料較薄時，在燒結過程中會出現裂紋。基于經驗知識，可以將機器視覺自動檢測技術集中到缺陷產生區域，明確檢測方向，有助于提高缺陷檢測的可靠性。
In order to detect the cracks and scratches of products by machine vision automatic detection technology, it is necessary to understand how the surface defects of products are generated, and clearly classify the possible shapes of these defects in the image. For example, when powder metal parts are thin, cracks will appear during sintering. Based on the experience knowledge, the machine vision automatic detection technology can be concentrated in the defect producing area, and the detection direction is clear, which is helpful to improve the reliability of defect detection.
在人工檢測過程中，通過直覺和經驗可以預測缺陷的形狀。例如，劃痕通常表現為一系列的直線或略微彎曲的線，這些線往往潛意識地通過心理上的“格式塔”能力連接成一條線。在機器視覺自動檢測技術中，這種“格式塔”能力可以用視覺微積分粗略地表現出來。
In the process of manual inspection, the shape of defects can be predicted by intuition and experience. For example, scratches usually appear as a series of straight lines or slightly curved lines, which are often subconsciously connected into a line by a psychological Gestalt ability. In machine vision automatic detection technology, this "Gestalt" ability can be roughly expressed by visual calculus.
在高難度測試過程中，測試人員可能會有不同的意見——哪是擦傷？或者只是碰巧形成一條線的亮點？這說明需要考慮缺陷檢測的可靠性。如果表面缺陷清晰，缺陷成分清晰，機器視覺自動檢測技術的檢測結果的一致性和可靠性高于人工視覺檢測。
In the course of a difficult test, testers may have different opinions - which is a scratch? Or is it just a bright spot that happens to form a line? This shows that the reliability of defect detection needs to be considered. If the surface defects are clear and the defect components are clear, the consistency and reliability of the detection results of machine vision automatic detection technology are higher than that of artificial vision inspection.
為了檢測產品表面上的小缺陷，確保零件表面始終處于相機的焦點區域是非常重要的。視覺自動檢測技術用于準確檢測產品的裂紋和劃痕，有關機器視覺自動檢測技術的更多信息，請訪問山東紅寶自動化有限公司官方網站:http://www.52iku.com。
In order to detect small defects on the product surface, it is very important to ensure that the part surface is always in the focus area of the camera. Visual automatic detection technology is used to accurately detect cracks and scratches on products. For more information about machine vision automatic detection technology, please visit the official website of Shandong Hongbao Automation Co., Ltd.
視覺自動檢測技術