Method and apparatus for differentiating a planar textured surface from a surrounding background

Claim 3

The method according to claim 1 further including the step of locating at least one of the edges of the pattern of interest in the binarized image.

Claim 4

The method according to claim 2 further including the step of inspecting the located edge for any cracks.

Claim 5

The method according to claim 2 wherein each edge of the pattern is located by determining, at preselected sampling intervals along one of a pair of coordinate axes, the first location where a large difference exists between the intensity of a pair of neighboring pixels within the binarized image and then mathematically defining the edge from the locations of the pixel intensity differences.

Claim 6

The method according to claim 4 wherein the edge is inspected for cracks by comparing, at each sampling interval, the location of the edge, as established by the location of the pixel intensity difference, to the location of the edge given by the mathematical definition thereof.

Claim 7

In combination with a process of manufacturing an electronic component by bonding a semiconductor chip within a recess in a chip carrier so that a major surface of the chip remains exposed, a method for inspecting the edges of the exposed surface of the chip to detect cracks, if any, comprising the steps of: illuminating both the exposed surface of the chip and the surrounding bonding material by directing light thereat so the light strikes the exposed surface of the chip and the surrounding bonding material substantially normal to the plane of the exposed chip surface; capturing the image of the exposed chip surface and the surrounding background bonding material with an image-acquisition device which is positioned to sense the light reflected normally from the background region and the exposed chip surface such that only the exposed chip surface appears in focus within the captured image; determining the intensity of each of a plurality of small areas (pixels) within the captured image; determining the intensity gradient for each pixel; creating a gradient image by assigning each pixel within the captured image an intensity value according to its gradient intensity; binarizing the gradient image by causing each pixel to appear bright or dark, depending on whether the intensity gradient established for each pixel is above or below, respectively, a predetermined threshold, thereby effectively isolating the chip within the captured image; establishing the expected location of each of the edges of the chip within the captured image; and inspecting each of the edges by detecting, at preselected intervals along each of a pair of coordinate axes, the coordinate location of where the edge of the chip is found within the image thereof and then determining whether the actual location of the edge of the chip substantially coincides with the expected location of the edge.

Claim 8

In combination with a process of manufacturing an electronic component by bonding a semiconductor chip within a recess in a chip carrier so that a major surface of the chip remains exposed, a method for inspecting the edges of the exposed surface of the chip to detect cracks, if any, comprising the steps of: illuminating both the exposed surface of the chip and the surrounding bonding material by directing light thereat so the light strikes the exposed surface of the chip and the surrounding bonding material substantially normal to the plane of the exposed chip surface; capturing the image of the exposed chip surface and the surrounding background bonding material with an image-acquisition device which is positioned to sense the light reflected normally from the background region and the exposed chip surface such that only the exposed chip surface appears in focus within the captured image; determining the intensity of each of a plurality of small areas (pixels) within the captured image; determining the intensity gradient for each pixel; creating a gradient image by assigning each pixel within the captured image an intensity value according to its gradient intensity; binarizing the gradient image by causing each pixel to appear bright or dark, depending on whether the intensity gradient established for each pixel is above or below, respectively, a predetermined threshold, thereby effectively isolating the chip within the captured image; establishing the expected location of each of the edges of the chip within the captured image; and inspecting each of the edges by detecting, at preselected intervals along each of a pair of coordinate axes, the coordinate location of where the edge of the chip is found within the image thereof and then determining whether the actual location of the edge of the chip substantially coincides with the expected location of the edge.

Claim 9

The method according to claim 6 wherein the actual location of the edges of the chip is determined from the locations, at preselected intervals along each of a pair of coordinate axes, of the first location where there is a significant difference in the intensity between a neighboring pair of pixels.

Claim 10

The method according to claim 6 wherein the expected location of the edges of the chip is established by first determining, at selected intervals along each of a pair of coordinate axes within the captured image, the locations of the first significant transitions in the pixel intensity, and then performing a linear regression analysis on the locations of the first pixel intensity transitions to establish a mathematical definition for each edge which predicts the location thereof within the image.

Claim 11

Apparatus for differentiating between a textured planar surface on an article and a background region surrounding the pattern, comprising: means for illuminating both the textured planar surface and the surrounding background region by directing light thereat so the light strikes the surface and the surrounding background region substantially normal to the plane of each; image-acquisition means for capturing the image of the surface and the surrounding background region, the image-acquisition means positioned such that its optical axis is substantially normal to the surface and the surface appears in focus whereas the background region appears out of focus within the captured image; means for determining the intensity of each of a plurality of small areas (pixels) within the captured image; means for determining the intensity gradient for each pixel; means for establishing a gradient image of the surface and the surrounding background region; and means for binarizing the gradient image by causing each pixel to appear bright or dark, depending on whether the gradient established for each pixel is above or below a predetermined threshold, respectively, thereby eliminating the background region from the captured image.

Claim 12

Apparatus for differentiating between a textured planar surface on an article and a background region surrounding the pattern, comprising: means for illuminating both the textured planar surface and the surrounding background region by directing light thereat so the light strikes the surface and the surrounding background region substantially normal to the plane of each; image-acquisition means for capturing the image of the surface and the surrounding background region, the image-acquisition means positioned such that its optical axis is substantially normal to the surface and the surface appears in focus whereas the background region appears out of focus within the captured image; means for determining the intensity of each of a plurality of small areas (pixels) within the captured image; means for determining the intensity gradient for each pixel; means for establishing a gradient image of the surface and the surrounding background region; and means for binarizing the gradient image by causing each pixel to appear bright or dark, depending on whether the gradient established for each pixel is above or below a predetermined threshold, respectively, thereby eliminating the background region from the captured image.

Claim 13

The apparatus according to claim 9 further including means for detecting the edges of the pattern of interest and thereby locate the chip.

Claim 14

The apparatus according to claim 10 further including means for inspecting the edges of the chip to detect edge cracks, if any.

Claim 15

The apparatus according to claim 10 further including means for matching the textured planar surface to that of a predetermined pattern.

Claim 16

The apparatus according to claim 9 wherein the means for determining the intensity of the pixels in the captured image, the means for establishing the intensity gradient, means for establishing the gradient image, and the means for binarizing the captured image to isolate the image of the textured surface comprise a machine vision processor.