JPS59163861A - Detector for white flaw and black flaw of solid-state image pickup element - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention provides a solid-state imaging device that can automatically and mechanically detect white flaws and black flaws on a solid-state imaging device according to predetermined standards without visual inspection. The present invention relates to a device for detecting white flaws and black flaws on an element.

[Background of the invention]

Conventionally, white and black scratches on solid-state imaging devices have been detected by shining light at a constant intensity onto the light-receiving surface of the device under test and visually observing the screen of a monitor television (by the inspector's eyes), or by similar methods. This was done by applying uniform light to the light-receiving surface of the device under test and comparing the entire surface at a fixed slice level (using an inspection machine).

However, when using the inspector's eyes, the Y1 constant standard is likely to vary and the measurement time is long.Furthermore, when a fixed slice level is used, unevenness in the light used for measurement, uneven sensitivity of the device under test, etc. can be mistaken for flaws. However, there were problems such as the inability to detect minute scratches.

[Object 1 of the Invention] [Object 1] The object of the present invention is to provide a detection device for IM that can automatically and mechanically detect white flaws and black flaws in a solid-state ilSl image without missing even the slightest flaw.

[Summary of the invention]

In order to achieve the above object, in the present invention, the output from each pixel when scanning is repeated multiple times is A/D converted and averaged, and the output average of multiple pixels in the vicinity of each pixel is calculated. By comparing the values, we detected defects that were too high or too low, and determined them to be white or black scratches, respectively.

The reason why the average value of pixel outputs generated during multiple repeated scans is used is that it is possible to avoid the effects of random noise that may cause erroneous determinations. The human eye also perceives the average value of repeated scanning outputs, and ignores random noise that occurs only once. In order to add up the outputs generated by repeatedly scanning each pixel forming the screen multiple times, it is sufficient to have a storage device that can store data corresponding to each pixel on both sides (frames).

Output of each pixel (of course, average value of two repeated outputs)
The reason why each pixel is compared with the average output of multiple pixels in the vicinity of each pixel is that it can be used even if there is unevenness in the light intensity irradiating the test element, or if the test element has gradual sensitivity fluctuations. This is also because by doing so, it is possible to prevent a situation in which a erroneous determination as <<(K) is made as in the conventional case using a fixed slice level.

It takes a considerable amount of time to add up the results of multiple scans, but compared to visual inspection by humans, the time required is much shorter, and it can be done automatically and mechanically. , the judgment accuracy is improved, unevenness is eliminated, and the number of inspection steps is reduced.

[Embodiments of the invention]

FIG. 1 shows a schematic block diagram of one embodiment of the present invention, 1
2 is a driving circuit, 2 is a solid-state image sensor to be tested, 3 is an amplifier, 4 is a to/D converter, 5 is a frame memory, 6 is a determination circuit,
7shi4. It is a valence counter. Drive circuit +4:i, generates clocks, power supplies, etc. necessary for operation of the solid-state image sensor 2 to be tested. The amplifier 3 converts the output of the pixel of the solid-state image sensor into A
/D convertible voltage level). The A/D converter 4 converts the pixel output into digital data to facilitate subsequent handling.4 bits can indicate the output stage of at least [V, and 6 bits is almost sufficient. The frame memory 5 is a memory that stores the output of each pixel of the solid-state image sensor 2, and stores the results of multiple additions to absorb random noise. The determination circuit 6 compares the contents stored in the frame memory with the slice level and determines the presence or absence of scratches. As mentioned above, in order to absorb the uneven illuminance of the light irradiating the device under test and the non-uniformity (sensitivity unevenness) of the output of the solid-state image sensor, an average is taken for each of a plurality of n pixels (samples) in the vicinity of the device, and this average value is Scratches are determined based on the difference from (slice level). FIG. 2 shows the state of judgment by this judgment circuit 6. As mentioned above, if the average value of the outputs of n pixels in the vicinity of each pixel is used as the judgment standard, there will be uneven illuminance on the surface of the element to be tested. It can be seen that there is no problem even if the level is gradually inclined. The valence counter 7 is a judgment circuit 6
The number of flaws detected is counted, and this number is used as data for quality determination.

〔Effect of the invention〕

As explained above, according to the present invention, accurate measurement can be performed without being affected by uneven illuminance of the device under test, uneven sensitivity of the device under test, and random noise that is generally difficult to avoid in electrical circuits. It becomes possible to automatically detect white flaws and black flaws on a solid-state image sensor while maintaining determination accuracy.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram of an embodiment of the present invention, and FIG. 2 is a diagram showing a determination state by a determination circuit according to the present invention. 1-drive circuit, 2-test solid-state image sensor, 3-amplifier, 4
---A/D converter, 5-frame memory, 6-
Judgment circuit, 7-valence counter. Agent Patent Attorney Takeo Agata

Claims (1)

[Claims] A solid-state image sensor that shines uniform light onto the solid-state image sensor, and detects the output from each pixel that is too high or too low due to this light and determines it as a white flaw or a black flaw. In a scratch/black scratch detection device, the output from each pixel is A/D converted and averaged when scanning is repeated multiple times, and the average value is compared with the average output value of multiple pixels in the vicinity of each pixel. A device for detecting white flaws and black flaws in a solid-state image sensor, characterized in that the flaws are detected when the flaws are too high or too low, and the flaws are determined to be white flaws or black flaws, respectively.