Sparse path codes and methods
A series of marks and absences of marks (voids) arrayed along one or a few mathematically-defined paths, define a message-conveying sparse path code. Multiple improvements in the forms of such codes, and related encoding and reading techniques, are detailed. Some such improvements provide greatly increased robustness and decreased visibility. A variety of other features and arrangements are also detailed.
1 . A plastic sorting method comprising the acts:
reading a sparse path code defined by features on an item, from imagery of said item on a conveyor and depicting the sparse path code, the sparse path code following a curved line and comprising regularly-spaced candidate marking locations along the curved line at which said features are selectively placed, the reading producing information, in which the reading act includes extracting a bit sequence from the sparse path code, and computing a dot product between an extracted bit sequence, and each of plural candidate sequences, to produce said information by determining which of said plural candidate sequences is most likely to correspond to the extracted bit sequence;
using produced information to determine that the item comprises an item of one plastic, sheathed in a sleeve of a second, different plastic; and
diverting said item into a repository that contains other sheathed items.
2 . The plastic sorting method of claim 1 in which said reading comprises applying the imagery to a network that includes one or more convolutional layers and that is configured to identify a location of at least a sync code portion of the sparse path code in the imagery.
3 . The method of claim 1 in which the reading act includes applying a Laplacian filter, or plural Laplacian filters of different scales, to the imagery depicting the sparse path code.
4 . The method of claim 1 in which the reading act includes examining the imagery depicting the sparse path code for pairs of features located within a predetermined pixel spacing.
5 . The method of claim 1 in which the reading act includes determining a probability value for each sparse path code feature depicted in the imagery, based on a local contrast metric computed over a region centered on the feature, and performing soft-decision decoding employing determined probability values for said features in producing said information.
6 . The method of claim 1 in which the sparse path code comprises a closed-path code.
7 . The method of claim 1 wherein one of said features spans an area on said item that is X microns on a side, and a pixel in said imagery spans a camera focal plane region that is Y microns on a side, wherein the ratio X/Y is less than 1, 1/2, 1/4, or 1/8.
8 . The method of claim 1 in which the features are 150 microns, 120 microns, 100 microns, 75 microns, 66 microns, 50 microns, 40 microns, 33 microns, or 20 microns or less in diameter.
9 . A system for sorting plastic items, comprising:
a conveyor having an item thereon;
a camera system for capturing imagery depicting the conveyor with the item thereon;
means for reading a sparse path code defined by features on the item, from captured imagery depicting the sparse path code on the item, the sparse path code following a curved line and comprising regularly-spaced candidate marking locations along the curved line at which said features are selectively placed, said means producing information, in which said means for reading a sparse path code includes means for extracting a bit sequence from the sparse path code and for computing a dot product between an extracted bit sequence, and each of plural candidate sequences, in producing the information by determining which of the plural candidate sequences is most likely to correspond to the extracted bit sequence;
means for using produced information to determine the item comprises an item of one plastic, sheathed in a sleeve of a second, different plastic; and
means for diverting said item into a repository that contains other sheathed items.
10 . The system of claim 9 in which said means for reading comprises a network including one or more convolutional layers, in which the network is configured to identify a location of at least a sync code portion of the sparse path code in the captured imagery.
11 . The system of claim 9 in which the means for reading a sparse path code includes means for applying a Laplacian filter, or plural Laplacian filters of different scales, to imagery depicting the sparse path code.
12 . The system of claim 9 in which the means for reading a sparse path code includes means for examining the imagery depicting the sparse path code for pairs of features located within a predetermined pixel spacing.
13 . The system of claim 9 in which the means for reading a sparse path code includes means for determining a probability value for each sparse path code feature depicted in the imagery, based on a local contrast metric computed over a region centered on the feature, and performing soft-decision decoding employing determined probability values for said features in producing said information.
14 . The system of claim 9 in which the sparse path code comprises a closed-path code.
15 . The system of claim 9 wherein one of said features spans an area on said item that is X microns on a side, and a pixel in said imagery spans a camera focal plane region that is Y microns on a side, wherein the ratio X/Y is less than 1, 1/2, 1/4, or 1/8.
16 . The system of claim 9 in which the features are 150 microns, 120 microns, 100 microns, 75 microns, 66 microns, 50 microns, 40 microns, 33 microns, or 20 microns or less in diameter.
17 . A non-transitory computer readable medium containing instructions for configuring a computer system in a plastic sorting facility to perform acts including:
reading a sparse path code defined by features on an item, from imagery of said item on a conveyor, the sparse path code following a curved line and comprising regularly-spaced candidate marking locations along the curved line at which said features are selectively placed, the reading producing information, in which the reading a sparse path code includes extracting a bit sequence from the sparse path code, and computing a dot product between an extracted bit sequence, and each of plural candidate sequences, to produce the information by determining which of said plural candidate sequences is most likely to correspond to the extracted bit sequence;
using produced information to determine that the item comprises an item of one plastic, sheathed in a sleeve of a second, different plastic; and
controlling a diverter apparatus to divert the item from the conveyor into a repository that contains other sheathed items.