X-axis spatial redundancy supression: Contribution to the integration of smart reading techniques in a standard CMOS vision sensor

Classical image processing is based on the evaluation of data delivered by a vision sensor in the form of images. These sensors with conventional clocks, receive visual information from the scene sequentially. Each array of data that is transmitted, recorded and that requires generally a post processing, carries the information of all pixels of the matrix without considering the constancy of some values. Therefore, this method leads us, depending on the dynamic content of the scene, to a high level of data redundancy. Note that in this case of sequential reading, the output bandwidth of the sensor is shared equally among all the pixels in the matrix. Thus, we chose to implement in our imager a new readout control architecture which reduces the dataflow that goes through the ADC and that must be transmitted all the way after the converter. This architecture is based on the principle of the suppression of x-axis spatial redundancy.