Test System for CMOS Digital Image Sensors

Overview

Dark Field defective pixels are characterized as pixels that are “hot” or lighter intensity than other pixels with the same exposure. Most CMOS image sensors have on-chip algorithms to correct bad pixels and defects, but the processing of these pixels is limited. Testing and limiting the number of bad or defective pixels will ensure that the algorithm can accurately compensate for the defects.

This test is performed by first : shielding the image sensor from all light, and then acquiring an image of the uniform dark field. The defective pixel maximum intensity can be configured for this test and every pixel with an intensity greater than the specified limit is deemed defective. The defective pixel locations are reported for each Bayer color plane.

A row or column average graph is also provided to allow quick identification of bad rows or columns. An example of a bad column can be seen in the peaks that rise significantly above the means of the other rows or columns.

.

Dark Field Dead and Defective Pixel Detection Algorithm

The Bayer Dark Field Dead and Defective Pixel detection algorithm is a simple thresholding operation that classifies any pixel with an intensity exceeding the specified limit as a defective pixel. The test should be performed with the sensor shielded from all light. These pixels also called “Hot Pixels”. The (X, Y) locations of all detected outliers are reported.

A standard 2D image display is output, along with a 3D display of the dark field pixel intensities.