Cross-talk rejecting convolution peak finding

1 . A method of processing a histogram generated by a time-of-flight (ToF) imager, the method comprising:

filtering the histogram using a zero-crossing filter (ZCF) to generate a ZCF output signal;

finding zero-crossing points in the ZCF output signal, wherein the zero-crossing points define one or more pulse regions in the ZCF output signal;

computing, for each pulse region of the one or more pulse regions, a weighted sum of the pulse region;

finding, in each pulse region, a maximum peak;

classifying the maximum peak in each pulse region as a first type of peak or a second type of peak based on the weighted sum of the pulse region; and

generating a list of ZCF targets from the maximum peaks classified as the first type of peaks.

2 . The method of claim 1 , wherein computing, for each pulse region of the one or more pulse regions, the weighted sum of the pulse region comprises:

multiplying values of the ZCF output signal in the pulse region with respective weight coefficients of the histogram to generate weighted ZCF output values; and

summing up the weighted ZCF output values to generate the weighted sum of the pulse region.

3 . The method of claim 2 , wherein a weight coefficient is assigned for each histogram bin of the histogram, wherein a first plurality of weight coefficients before a reference zero-point of the histogram have negative values, and a second plurality of weight coefficients after the reference zero-point of the histogram have positive values.

4 . The method of claim 3 , wherein the reference zero-point of the histogram is a histogram bin that corresponds to a distance between an emitter of the ToF imager and a window of an assembly housing of the ToF imager, wherein the emitter is configured to send a light signal through the window of the assembly housing.

5 . The method of claim 2 , wherein classifying the maximum peak in each pulse region comprises:

classifying the maximum peak as the first type of peak if the weighted sum of the pulse region is larger than a positive threshold or smaller than a negative threshold; and

classifying the maximum peak as the second type of peak if the weighted sum of the pulse region is between the positive threshold and the negative threshold.

6 . The method of claim 5 , further comprising:

filtering the histogram using a matched filter (MF) to generate an MF output signal;

finding a strongest MF target in the MF output signal; and

generating a list of targets in the histogram by combining the strongest MF target and the list of ZCF targets.

7 . The method of claim 6 , further comprising:

identifying the strongest MF target as a strongest target in the histogram if an amplitude of the strongest MF target is larger than an amplitude of a strongest ZCF target in the list of ZCF targets, or if the list of ZCF targets is empty.

8 . The method of claim 1 , wherein generating the list of ZCF targets comprises:

comparing a confidence level of each of the maximum peaks classified as the first type of peak with a confidence threshold; and

copying the maximum peaks classified as the first type of peak and with confidence levels higher than the confidence threshold to the list of ZCF targets.

9 . The method of claim 1 , wherein the ZCF comprises:

a ZCF pulse region, wherein a width of the ZCF pulse region corresponds to a full width at half maximum (FWHM) of a pulse shape of an emitter of the ToF imager;

a pre-pulse region before the ZCF pulse region, wherein first filter coefficients in the pre-pulse region have negative values; and

a post-pulse region after the ZCF pulse region, wherein second filter coefficients in the post-pulse region have negative values, wherein a second average of the second filter coefficients is smaller than a first average of the first filter coefficients.

10 . The method of claim 9 , wherein a sum of the coefficients of the ZCF is a negative value.

11 . A method of processing a histogram generated by a time-of-flight (ToF) imager, the method comprising:

assigning a weight coefficient for each histogram bin of the histogram;

generating a ZCF output signal by filtering the histogram using a zero-crossing filter (ZCF);

finding positive pulse regions in the ZCF output signal, wherein each of the positive pulse regions is defined by a pair of zero-crossing points;

computing, for each positive pulse region, a weighted sum by multiplying values of the ZCF output signal with corresponding weight coefficients of the histogram bins to generate weighted values of the ZCF output signal and by adding the weighted values of the ZCF output signal in each positive pulse region;

classifying a maximum peak in each positive pulse region as a first type of peak or a second type of peak based on the weighted sum of the positive pulse region; and

creating a list of ZCF targets from the maximum peaks classified as the first type of peaks.

12 . The method of claim 11 , further comprising:

identifying a strongest ZCF target from the list of ZCF targets;

generating a matched filter (MF) output signal by filtering the histogram with an MF;

identifying a strongest MF target from the MF output signal; and

choosing, from the strongest ZCF target and the strongest MF target, a stronger target as a strongest target in the histogram.

13 . The method of claim 11 , wherein assigning the weight coefficient comprises:

assigning negative values for the weight coefficients of histogram bins located before a pre-determined histogram bin; and

assigning positive values for the weight coefficients of histogram bins located after the pre-determined histogram bin.

14 . The method of claim 13 , wherein the weight coefficients of the histogram bins increase along a direction from a first histogram bin of the histogram toward a last histogram bin of the histogram.

15 . The method of claim 13 , wherein classifying the maximum peak in each positive pulse region comprises:

classifying the maximum peak as the first type of peak if the weighted sum of the positive pulse region is larger than a first threshold or smaller than a second threshold, wherein the first threshold is larger than the second threshold; and

classifying the maximum peak as the second type of peak if the weighted sum of the positive pulse region is between the second threshold and the first threshold.

16 . The method of claim 15 , wherein the first threshold is a positive value, and the second threshold is a negative value.

17 . A time-of-flight (ToF) imager comprising:

a light source configured to generate a light signal for illuminating an object;

a single-photon avalanche diode (SPAD) configured to generate a histogram for a reflected light signal from the object;

a zero-crossing filter (ZCF) configured to generate a ZCF output signal by filtering the histogram;

a matched filter (MF) configured to generate an MF output signal by filtering the histogram; and

a peak finding circuit configured to identify a list of targets in the histogram using the ZCF output signal and the MF output signal by:

finding positive pulse regions in the ZCF output signal, wherein each of the positive pulse regions is defined by a pair of zero-crossing points;

computing, for each positive pulse region, a weighted sum of the ZCF output signal in the positive pulse region;

classifying a maximum peak in each positive pulse region as a first type of peak or a second type of peak based on the weighted sum of the positive pulse region;

generating a list of ZCF targets based on the maximum peaks classified as the first type of peaks;

identifying a strongest MF target from the MF output signal; and

generating the list of targets in the histogram using the list of ZCF targets and the strongest MF target.

18 . The ToF imager of claim 17 , wherein the peak finding circuit is configured to identify a strongest target in the histogram by:

identifying a stronger one of the strongest MF target and a strongest ZCF target in the list of ZCF targets as the strongest target in the histogram.

19 . The ToF imager of claim 17 , wherein the peak finding circuit is configured to compute, for each positive pulse region, the weighted sum of the ZCF output signal in the positive pulse region by:

multiplying values of the ZCF output signal in the positive pulse region with corresponding weight coefficients to generate weighted values of the ZCF output signal; and

adding up the weighted values of the ZCF output signal in the positive pulse region.

20 . The ToF imager of claim 19 , wherein the peak finding circuit is configured to classify the maximum peak in each positive pulse region by:

classifying the maximum peak as the first type of peak if the weighted sum of the positive pulse region is larger than a first threshold or smaller than a second threshold, wherein the first threshold is larger than the second threshold; and

classifying the maximum peak as the second type of peak if the weighted sum of the positive pulse region is between the second threshold and the first threshold.