# US 12,214,601 B2 · Detecting a defective nozzle in a digital printing system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,214,601 B2 - Document ID: US12214601B2 - Status: Granted - Kind: B2 - Application number: 17/921,359 - Filed: 2021-05-12 - Published: 2023-08-24 - Granted: 2025-02-04 - Earliest claimed: 2020-05-17 - Assignees: Landa Corporation Ltd. - Applicants: Landa Corporation Ltd. - Inventors: Boris Levant; Shai Silberstein; Tomer Yanir; Avraham Guttman; Alon Siman Tov - Web page: https://patentgrep.com/patent/US12214601B2 ## CPC - B41J 2/2142: {Detection of malfunctioning nozzles (generating single droplets or particles on demand by pressure, e.g. electromechanical transducers B41J2/045, B41J2/05; jet deflection sensors B41J2/125; for cleaning purposes B41J2/16579)} - B41J 2/2146: {for line print heads} - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - G06T 7/0004: {Industrial image inspection} - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - B41J: TYPEWRITERS; SELECTIVE PRINTING MECHANISMS, i.e. MECHANISMS PRINTING OTHERWISE THAN FROM A FORME; CORRECTION OF TYPOGRAPHICAL ERRORS - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A method includes, receiving a first digital image (FDI) to-be printed by a digital printing system (DPS) (10). In a in training phase: for first selected regions (111) in the FDI, a first set of synthetic images (SIs) (112A, 112B, 114A, 114B, 116A, 116B) having a defect caused by a defective part (DP) (99) in the first selected regions, is produced; a neural network (NN) (150) is trained to detect the defect using the first set SIs. In a subsequent detection phase: the NN is applied for identifying, in a second digital image (SDI) (136, 146) acquired from an image produced by the DPS, suspected second regions (135, 145); for each of the second regions, a second set (137, 147) of SIs having DPs that form the defects, is produced; and the DP is identified by comparing, in each of the second regions, between the SDI and the second set SIs. ## Claims ### Claim 1 A method for detecting a defective part (DP) in a digital printing system (DPS), the method comprising: receiving a first digital image (FDI) to be printed by the DPS; in a training phase: producing, for one or more first selected regions in the FDI, a first set of one or more synthetic images having a defect caused by the DP in the one or more first selected regions; selecting, based on a predefined selection criterion, the first selected regions that comprise features for training a neural network (NN); and training the NN to detect the defect using at least one of the synthetic images of the first set; and in a detection phase that is subsequent to the training phase: applying the trained NN for identifying, in a second digital image (SDI) acquired from a printed image produced by the DPS, one or more second regions suspected of having the defect; producing, for each of the second regions, a second set of one or more synthetic images having one or more DPs producing respectively one or more of the defects; and identifying at least the DP by comparing, in each of the second regions, between the SDI and the one or more synthetic images of the second set. ### Claim 2 (depends on 1) The method according to claim 1, wherein the NN comprises a convolutional NN (CNN). ### Claim 3 (depends on 2) The method according to claim 2, wherein the CNN has an inception-v3 architecture. ### Claim 4 (depends on 1) The method according to claim 1, wherein at least one of the FDI and SDI comprises a product image. ### Claim 5 (depends on 1) The method according to claim 1, wherein the DPS comprise nozzles for directing a printing fluid onto a substrate, wherein the DP comprises a defective nozzle (DN) from among the nozzles, and wherein the defect comprises a missing nozzle fault (MNF) caused by a blocked orifice of the DN. ### Claim 6 (depends on 1) The method according to claim 1, wherein the DPS comprise nozzles for directing a printing fluid onto a substrate, wherein the DP comprises a partially-clogged nozzle from among the nozzles, and wherein the defect comprises a registration error caused by the partially-clogged nozzle, which directs a printing fluid jetted at a deflected angle to land on a substrate at a distance from an intended landing position. ### Claim 7 A system for detecting a defective part (DP) in a digital printing system (DPS), the system comprising: an interface, which is configured to receive: (i) a first digital image (FDI) to be printed by the DPS, and (ii) a second digital image (SDI) acquired from a printed image produced by the DPS; and a processor, wherein in a training phase, the processor is configured, to: (i) produce, for one or more first selected regions in the FDI, a first set of one or more synthetic images having a defect caused by the DP in the one or more first selected regions, (ii) select, based on a predefined selection criterion, the first selected regions that comprise features for training a neural network (NN), and (iii) train the NN to detect the defect using at least one of the synthetic images of the first set; and in a detection phase that is subsequent to the training phase, the processor is configured, to: (i) apply the trained NN for identifying, in the SDI, one or more second regions suspected of having the defect, (ii) produce, for each of the second regions, a second set of one or more synthetic images having one or more DPs producing respectively one or more of the defects, and (iii) identify at least the DP by comparing, in each of the second regions, between the SDI and the one or more synthetic images of the second set. ### Claim 8 (depends on 7) The system according to claim 7, wherein the NN comprises a convolutional NN (CNN). ### Claim 9 (depends on 8) The system according to claim 8, wherein the CNN has an inception-v3 architecture. ### Claim 10 (depends on 7) The system according to claim 7, wherein at least one of the FDI and SDI comprises a product image. ### Claim 11 (depends on 7) The system according to claim 7, wherein the DPS comprise nozzles for directing a printing fluid onto a substrate, wherein the DP comprises a defective nozzle (DN) from among the nozzles, and wherein the defect detected by the NN comprises a missing nozzle fault (MNF) caused by a blocked orifice of the DN. ### Claim 12 (depends on 7) The system according to claim 7, wherein the DPS comprise nozzles for directing a printing fluid onto a substrate, wherein the DP comprises a partially-clogged nozzle from among the nozzles, and wherein the defect comprises a registration error caused by the partially-clogged nozzle, which directs a printing fluid jetted at a deflected angle to land on a substrate at a distance from an intended landing position.