公开(公告)号：US20230215034A1

公开(公告)日：2023-07-06

申请号：US17894250

申请日：2022-08-24

Applicant: CANON KABUSHIKI KAISHA

Inventor： Tatsuya Yamazaki

IPC: G06T7/70 ,  G06V40/16 ,  G06V40/18 ,  G06T7/20 ,  G06T7/62

CPC classification number: G06T7/70 ,  G06V40/161 ,  G06V40/18 ,  G06T7/20 ,  G06T7/62 ,  G06T2207/30201 ,  G06T2207/30248 ,  G06V2201/07

Abstract: An image processing apparatus detects a subject of a first type and a subject of a second type to an image. When executing tracking processing of a subject based on a detection result of the detection circuit, if a same subject is detected as a subject of the first type and a subject of the second type, the image processing apparatus selects either the detection result regarding the subject of the first type or the detection result regarding the subject of the second type is to be used to perform the tracking processing of the subject.

公开(公告)号：US20230209188A1

公开(公告)日：2023-06-29

申请号：US18067034

申请日：2022-12-16

Applicant: CANON KABUSHIKI KAISHA

Inventor： Akihiko KANDA ,  Hiroshi Yashima ,  Hideki Ogura ,  Hideyuki Hamano

IPC: H04N23/67 ,  G06T7/70 ,  G06V10/74

CPC classification number: H04N23/675 ,  G06T7/70 ,  G06V10/761 ,  G06V2201/07 ,  G06T2207/20044 ,  G06T2207/30196

Abstract: An image processing apparatus detects, from an image, a subject(s) of a first type and a subject(s) of a second type. The apparatus further detects a posture for each of the subject(s) of the first type. The apparatus then obtains, for each of the subject(s) of the first type, reliability that the subject is a main subject, based on the posture, and obtains a focus condition for each of the subject(s) of the first type and each of the subject(s) of the second type. The apparatus determines, based on the reliability and the focus condition, a main subject from the subject(s) of the first type and the subject(s) of the second type detected from the image.

公开(公告)号：US20230206649A1

公开(公告)日：2023-06-29

申请号：US18089712

申请日：2022-12-28

Applicant: THINKWARE CORPORATION

Inventor： Shinhyoung Kim

IPC: G06V20/58 ,  G06V10/25 ,  G06T7/60 ,  B60W60/00 ,  B60W30/09 ,  B60W30/095

CPC classification number: G06V20/58 ,  G06V10/25 ,  G06T7/60 ,  B60W60/0015 ,  B60W30/09 ,  B60W30/0956 ,  B60W30/0953 ,  G06V2201/08 ,  G06T2207/30252 ,  G06V2201/07 ,  B60W2554/404 ,  B60W2520/00 ,  B60W2420/42 ,  B60W2554/80

Abstract: An electronic device provided in an autonomous vehicle, the electronic device comprising a camera, a memory storing at least one instruction, and at least one processor operatively coupled with the camera, wherein the at least one processor is configured to, when the at least one instruction is executed obtain a front image in which the autonomous vehicle is driving through the camera, identify a target vehicle in the front image based on the vehicle detection model stored in the memory, generate a bounding box corresponding to the target vehicle in response to an identification of the target vehicle, generate a sliding window having a height equal to the height of the bounding box and having a width half of the width of the bounding box, divide the bounding box into a first area positioned left based on a middle position of the width of the sliding window and a second area positioned right based on the middle position, generate an extended bounding box by extending the first area in a left direction and extending the second area in a right direction, wherein size of the extended bounding box is twice as wide as size of the bounding box, obtain a sum of a pixel difference values between the first area and the second area for each shift by sequentially shifting the sliding window by a predefined pixel interval with respect to all of width of the extended bounding box, and identify a point that corresponds to a minimum value among sum values respectively indicating the sums that are obtained according to the shifting.

公开(公告)号：US20230206573A1

公开(公告)日：2023-06-29

申请号：US18146805

申请日：2022-12-27

Applicant: VIRNECT inc.

Inventor： Ki Young Kim ,  Thorsten Korpitsch

IPC: G06T19/00 ,  G06V10/764

CPC classification number: G06T19/006 ,  G06V10/764 ,  G06V2201/07

Abstract: A method of learning a target object by detecting an edge from a digital model of the target object and setting a sample point according to one embodiment of the present disclosure, which is performed by a computer-aided design program of an authoring computing device, includes: displaying a digital model of a target object that is a target of image recognition; detecting edges on the digital model of the target object; classifying the detected edges according to a plurality of characteristics; obtaining sample point information on the detected edges; and generating object recognition library data for recognizing a real object implementing the digital model of the target object based on the detected edges, characteristic information of the detected edges, and the sample point information.

公开(公告)号：US20230206484A1

公开(公告)日：2023-06-29

申请号：US18145907

申请日：2022-12-23

Applicant: Jio Platforms Limited

Inventor： Tejas Sudam GAIKWAD ,  Bhupendra SINHA ,  Gaurav DUGGAL ,  Manoj Kumar GARG

IPC: G06T7/70 ,  G06T7/50 ,  G06T7/60 ,  G06V10/74

CPC classification number: G06T7/70 ,  G06T7/50 ,  G06T7/60 ,  G06V10/761 ,  G06V2201/07 ,  G06T2207/10024

Abstract: The present disclosure provides system and method for object detection in a discontinuous space. The system receives at least one captured image from one or more computing devices associated with one or more users. The at least one captured image comprises one or more objects in the discontinuous space, and the one or more objects are associated with at least one attribute. The system computes a score corresponding to each of at least one attribute of the one or more objects. The system detects the one or more objects in the discontinuous space based on the computed score. Further, the system determines a similarity grade for the one or more detected objects, where the similarity grade corresponds to an accuracy of inference for the one or more detected objects. Finally, the system updates a database based on the accuracy of inference to facilitate object detection in the discontinuous space.

公开(公告)号：US20230196797A1

公开(公告)日：2023-06-22

申请号：US17933366

申请日：2022-09-19

Applicant: Covidien LP

Inventor： Paul S. ADDISON ,  Michael ADDISON ,  Dean MONTGOMERY

IPC: G06V20/59 ,  G06T7/50 ,  G06T7/70 ,  G06V40/16

CPC classification number: G06V20/59 ,  G06T7/50 ,  G06T7/70 ,  G06V40/172 ,  G06T2207/30196 ,  G06V2201/07

Abstract: System and methods for non-contact monitoring in vehicles are described. The systems and methods may use depth sensing cameras as part of the non-contact monitoring system. In some embodiments, depth data from at least one depth sensing device that has a field of view of at least part of the interior of the vehicle is received, wherein the depth data represents depth information as a function of position across the field of view. The depth data is then processed to obtain further information related to the occupant within the vehicle.

公开(公告)号：US20230196722A1

公开(公告)日：2023-06-22

申请号：US18066356

申请日：2022-12-15

Applicant: CANON KABUSHIKI KAISHA

Inventor： Hiroshi YOSHIKAWA ,  Takamasa TSUNODA

IPC: G06V10/75 ,  G06T7/11 ,  G06V10/25

CPC classification number: G06V10/758 ,  G06T7/11 ,  G06V10/25 ,  G06V2201/07 ,  G06T2207/20081

Abstract: A learning apparatus trains an estimator that executes a recognition task. The learning apparatus comprises: an obtaining unit configured to obtain a plurality of training data items including input data and supervisory data corresponding to the input data; a calculating unit configured to calculate statistic information relating to a predetermined perspective in the plurality of training data items; a determining unit configured to determine a degree of importance of each training data item included in the plurality of training data items based on the statistic information; and a control unit configured to control training of the estimator based on the degree of importance. The determining unit determines the degree of importance of each training data item such that unevenness of the plurality of training data items with respect to the predetermined perspective is reduced.

公开(公告)号：US11682203B2

公开(公告)日：2023-06-20

申请号：US17869879

申请日：2022-07-21

Applicant: Institute of Facility Agriculture, Guangdong Academy of Agricultural Science ,  Guangdong Laboratory for Lingnan Modern Agriculture

Inventor： Sai Xu ,  Huazhong Lu ,  Xin Liang

IPC: G06V20/10 ,  G06V10/77 ,  G06V10/74 ,  G06V10/141

CPC classification number: G06V20/188 ,  G06V10/141 ,  G06V10/761 ,  G06V10/7715 ,  G06V2201/07

Abstract: A feature extraction method of fruit spectrum includes taking a vector of each wavelength point in spectrum of samples as source data, and acquiring a sorting of all vectors by processing the source data by SPA; according to the sorting of the vectors, acquiring distribution points of each sample on a coordinate system; acquiring classification results of the samples by destructive analysis, and acquiring a number of first sample categories; acquiring a first Euclidean distance between the first sample categories; according to a sorting of the wavelength points, acquiring distribution points of each sample on the coordinate system; acquiring a number of second sample categories; acquiring a second Euclidean distance between the second sample categories; determining whether the first Euclidean distance is less than the second Euclidean distance; determine a (M+2)-th vector to be valid or invalid based on a comparison result.

公开(公告)号：US11681046B2

公开(公告)日：2023-06-20

申请号：US17508482

申请日：2021-10-22

Applicant: Zoox, Inc.

Inventor： Thomas Oscar Dudzik ,  Kratarth Goel ,  Praveen Srinivasan ,  Sarah Tariq

IPC: G01S17/86 ,  G06T3/00 ,  G06T7/593 ,  G01S7/48 ,  G01S17/931 ,  G06V20/58 ,  G06F18/214 ,  G06F18/25 ,  G06V10/774 ,  G06V10/776

CPC classification number: G01S17/86 ,  G01S7/4808 ,  G01S17/931 ,  G06F18/2155 ,  G06F18/254 ,  G06T3/0093 ,  G06T7/593 ,  G06V10/776 ,  G06V10/7753 ,  G06V20/58 ,  G06T2207/10012 ,  G06T2207/20081 ,  G06V2201/07

Abstract: Techniques for training a machine learned (ML) model to determine depth data based on image data are discussed herein. Training can use stereo image data and depth data (e.g., lidar data). A first (e.g., left) image can be input to a ML model, which can output predicted disparity and/or depth data. The predicted disparity data can be used with second image data (e.g., a right image) to reconstruct the first image. Differences between the first and reconstructed images can be used to determine a loss. Losses may include pixel, smoothing, structural similarity, and/or consistency losses. Further, differences between the depth data and the predicted depth data and/or differences between the predicted disparity data and the predicted depth data can be determined, and the ML model can be trained based on the various losses. Thus, the techniques can use self-supervised training and supervised training to train a ML model.

公开(公告)号：US20230186736A1

公开(公告)日：2023-06-15

申请号：US17924738

申请日：2020-05-20

Applicant: NEC Corporation

Inventor： Yu NABETO ,  Soma SHIRAISHI ,  Takami SATO ,  Katsumi KIKUCHI

IPC: G07G1/00 ,  G06V10/74 ,  G06Q20/20

CPC classification number: G07G1/0045 ,  G06V10/761 ,  G06Q20/208 ,  G06V2201/07

Abstract: An image acquisition unit (11) acquires a recognition processing image. A recognition unit (12) recognizes a product in the recognition processing image based on an estimation model. A registration unit (13) registers a result of the recognition in recognized product information. An output unit (14) outputs a result of the recognition. A correction reception unit (16) receives an input for correcting a result of the recognition. A correction unit (17) changes a result of the recognition to a result of the recognition after a correction, and also stores correction information in which a result of the recognition after a correction and the recognition processing image are associated with each other. A learning unit (18) performs relearning by using the recognition processing image stored as the correction information and updates the estimation model, when a number of the recognition processing image stored exceeds a predetermined value.