-
公开(公告)号:WO2020186779A1
公开(公告)日:2020-09-24
申请号:PCT/CN2019/117377
申请日:2019-11-12
申请人: 平安科技(深圳)有限公司
摘要: 一种图片信息识别方法,包括:接收第一终端上传的业务图片;当业务图片包含图表信息时,确定业务图片对应的图表类型;若图表类型为第一类型,提取业务图片中的图表线条,对多个图表线条进行拼接,得到第一图表;第一图表包括多个空白格;识别每个空白格对应的信息文本;将第一图表转换为第二图表;第二图表包括多个标准格;确定标准格与空白格之间的对应关系,根据对应关系将信息文本填充至标准格中,得到业务图片对应的业务图表,将业务图表存储至图表库。
-
公开(公告)号:WO2020181834A1
公开(公告)日:2020-09-17
申请号:PCT/CN2019/121053
申请日:2019-11-26
申请人: 北京三快在线科技有限公司
发明人: 洪天琦
IPC分类号: G06K9/34
摘要: 一种识别卡号的方法和装置、存储介质及电子设备。所述方法包括:获取卡号序列的字符位的分布格式信息,其中,所述分布格式信息包括所述卡号序列的字符位间距信息(S11);通过预先训练的神经网络模型识别目标图像中的字符序列,并获取识别出的所述字符序列的字符位间距信息(S12);判断识别出的所述字符序列中的字符位间距信息与获取的所述分布格式信息中的字符位间距信息是否一致(S13);若所述字符序列中的所述字符位间距信息与获取的所述分布格式信息中的所述字符位间距信息一致,确定识别出的所述字符序列为目标卡号(S14)。
-
公开(公告)号:WO2020173056A1
公开(公告)日:2020-09-03
申请号:PCT/CN2019/102027
申请日:2019-08-22
申请人: 百度在线网络技术(北京)有限公司
摘要: 本申请实施例公开了一种交通图像识别方法、装置、设备和介质。其中,方法包括:获取车辆采集的视频流并提取出视频流中的每帧图像作为第一图片;将第一图片输入至去干扰自编码器进行预处理,以过滤第一图片中的干扰,输出第二图片,其中,去干扰自编码器通过至少两类干扰样本集进行训练而得,不同类型干扰样本集中所加入的扰动方式包括下述至少两种:噪音、仿射变化、滤波模糊化、亮度变化和单色化;将第二图片输入至交通标志识别模型来进行识别处理。本申请实施例提供了正确识别对抗样本对交通标志识别模型进行攻击后的交通标志的方式;可以降低交通图像中对抗样本的干扰,提高图像的识别正确率,提升无人驾驶或智能驾驶的安全性。
-
公开(公告)号:WO2020157157A1
公开(公告)日:2020-08-06
申请号:PCT/EP2020/052200
申请日:2020-01-29
申请人: PROPHESEE
发明人: LAGORCE, Xavier
摘要: The event-based sensor has an array of pixels facing a scene. The information to be processed includes events originating from the pixels depending on variations of incident light from the scene. The method comprises: storing (S1) a data structure for a set of pixels of the array, the data structure including, for each pixel of the set, event data associated with a most recent event originating from said pixel; and, upon receiving a current event (E) from a first pixel of the array, retrieving (S2) any event data included in the data structure for a group of pixels including the first pixel and a plurality of second pixels adjacent to the first pixel in the array; and labelling (S3) the current event based on at least one connectivity criterion between the current event (E) and the most recent event originating from a pixel of said group.
-
公开(公告)号:WO2020146133A1
公开(公告)日:2020-07-16
申请号:PCT/US2019/068213
申请日:2019-12-23
发明人: WOOD, Erroll William , BLEYER, Michael , EDMONDS, Christopher Douglas , FENTON, Michael Scott , FINOCCHIO, Mark James , JUDNICH, John Albert
摘要: A method to identify one or more depth-image segments that correspond to a predetermined object type is enacted in a depth-imaging controller operatively coupled to an optical time-of-flight (ToF) camera; it comprises: receiving depth-image data from the optical ToF camera, the depth-image data exhibiting an aliasing uncertainty, such that a coordinate (X, Y) of the depth-image data maps to a periodic series of depth values {Zk}; and labeling, as corresponding to the object type, one or more coordinates of the depth- image data exhibiting the aliasing uncertainty.
-
公开(公告)号:WO2020104308A1
公开(公告)日:2020-05-28
申请号:PCT/EP2019/081450
申请日:2019-11-15
摘要: Various embodiments of the present disclosure include a thermal ablation probabilistic controller (30) employing an ablation probability model (32) trained to render a pixel ablation probability for each pixel of an ablation scan image illustrative of a static anatomical ablation. In operation, the thermal ablation probabilistic controller (30) spatially aligns a temporal sequence of ablation scan datasets representative of a dynamic anatomical ablation, and applies the ablation probability model (32) to the spatial alignment of the temporal sequence of ablation scan datasets to render the pixel ablation probability for each pixel of the ablation scan image illustrative of the static anatomical ablation.
-
公开(公告)号:WO2020097909A1
公开(公告)日:2020-05-22
申请号:PCT/CN2018/115874
申请日:2018-11-16
申请人: 北京比特大陆科技有限公司
IPC分类号: G06K9/34
摘要: 本发明提供一种文本检测方法、装置及存储介质,通过神经网络模型获取目标图像中包括文字区域的掩码图;基于掩码图获取文字区域的第一检测框;若第一检测框满足预设切割条件,则对第一检测框进行切割,得到第二检测框;将第二检测框对应的图像作为文本检测结果。上述文本检测方法可用于处理长文本框以及曲形文本框,提高了获取文本检测框的精度。
-
38.
公开(公告)号:WO2020095130A1
公开(公告)日:2020-05-14
申请号:PCT/IB2019/058623
申请日:2019-10-10
申请人: SONY CORPORATION
IPC分类号: H04N5/232 , G06K9/00 , G06K9/32 , G06K9/34 , G06T5/00 , G06T5/50 , G06T7/571 , H04N5/262 , H04N5/265 , H04N5/272
摘要: An image processing device is disclosed. The image processing device controls an image capture device to capture at least first image that comprises one or more objects. The image processing device identifies at least a first object in the first image. The image processing device determines a first blur value of identified first object and determines the identified first object as a blur object based on the first blur value. The image processing device adjusts a focal point of the image capture device to focus on the blurred first object. The image processing device controls the image capture device to capture a second image of the first object based on adjusted focal point. The image processing device replaces the first object in first image with the first object in the second image to generate an output image and further controls the display screen to display well-focused output image.
-
公开(公告)号:WO2020043057A1
公开(公告)日:2020-03-05
申请号:PCT/CN2019/102587
申请日:2019-08-26
申请人: 腾讯科技(深圳)有限公司
发明人: 辛遥
摘要: 本申请涉及一种图片处理方法、任务数据处理方法和装置,所述方法包括:通过机器学习模型中各子结构所对应的处理单元执行相应的子任务来实现,至少部分所述处理单元包括FPGA单元;获取待处理图片;提取所述待处理图片中的文本特征;根据所述文本特征确定所述待处理图片中任意角度的候选文本框;对各所述候选文本框进行旋转感兴趣区域的池化处理,将各所述候选文本框投影到固定大小的特征图,得到各所述候选文本框对应的文本框特征图;识别所述文本框特征图中文本,得到文本识别结果。本申请通过FPGA架构,可并行处理数据以实现上述图片处理方法,能够在降低成本和功耗的同时,提高待处理图片中文本识别准确率和效率。
-
公开(公告)号:WO2019185245A3
公开(公告)日:2019-10-03
申请号:PCT/EP2019/054229
申请日:2019-02-20
申请人: I.R.I.S.
发明人: COLLET, Frédéric , HAUTOT, Jordi , DAUW, Michel
摘要: An image processing system and an image processing method for localising recognised characters in an image. An estimation means is configured to estimate a first location of a recognised character that has been obtained by performing character recognition of the image. A determination means is configured to determine second locations of a plurality of connected components in the image. A comparison means is configured to compare the first location and the second locations, to identify a connected component associated with the recognised character. An association means is configured to associate the recognised character, the identified connected component, and the second location of the identified connected component.
-
-
-
-
-
-
-
-
-