WO2015032267A1 - 图像存储方法及其装置 - Google Patents

图像存储方法及其装置 Download PDF

Info

Publication number
WO2015032267A1
WO2015032267A1 PCT/CN2014/084294 CN2014084294W WO2015032267A1 WO 2015032267 A1 WO2015032267 A1 WO 2015032267A1 CN 2014084294 W CN2014084294 W CN 2014084294W WO 2015032267 A1 WO2015032267 A1 WO 2015032267A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
region
target area
predetermined period
interest
Prior art date
Application number
PCT/CN2014/084294
Other languages
English (en)
French (fr)
Inventor
罗兵华
朱江
浦世亮
Original Assignee
杭州海康威视数字技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 杭州海康威视数字技术股份有限公司 filed Critical 杭州海康威视数字技术股份有限公司
Priority to EP14841969.0A priority Critical patent/EP3043555B1/en
Priority to US14/916,893 priority patent/US9870509B2/en
Publication of WO2015032267A1 publication Critical patent/WO2015032267A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N9/00Details of colour television systems
    • H04N9/79Processing of colour television signals in connection with recording
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/50Context or environment of the image
    • G06V20/52Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/25Determination of region of interest [ROI] or a volume of interest [VOI]
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/96Management of image or video recognition tasks

Definitions

  • the present invention relates to the field of data storage, and in particular, to an image storage technology.
  • a security bayonet refers to a road traffic on-site monitoring of all motor vehicles passing through the bayonet point, such as a toll booth, a traffic station or a security checkpoint, etc. system.
  • the HD bayonet system plays an important role in applications such as law and order, criminal investigation, combat and crime prevention, and handling of sudden incidents. Especially in criminal investigation applications, it is necessary to find clues based on stored bayonet images. This method is already criminal investigation. An important technical means to solve the case.
  • criminal investigation may require viewing historical images of three months or more, which requires the security bayonet image to be stored for a long time. Now, the HD bayonet construction is more and more, and the large number of bayonet images need to be saved.
  • the current problem facing the HD security bayonet system is: The problem of long-term storage of large number of bayonet images.
  • Chinese patent CN201210568309 is the closest technical solution of the present invention, which discloses a method for reducing storage space.
  • the invention proposes a method for reducing image storage space, which solves the problem that the monitoring system in the prior art requires a large storage space to store image signals.
  • the technical solution is: acquiring image signals through a camera, buffering image signals, outputting data signals, compressing data signals, extracting GIF (Graphics Interchange Format) or JPG (Joint Photographic Experts Group, joint image)
  • GIF Graphics Interchange Format
  • JPG Joint Photographic Experts Group, joint image
  • the image data of the JPG format is extracted, otherwise the image data of the GIF format is extracted.
  • the essence of the technical solution is: in the case that there is no difference in the acquired images, the low-pixel images are stored in the memory, which reduces the storage space. This method of reducing storage space by reducing the resolution of the image, although reducing the storage space, also reduces the recognizability of key information, the loss of such key information will seriously affect the image of the security bayonet in criminal investigation.
  • an embodiment of the present invention discloses an image storage method, including the steps of: determining whether a storage time of a current image reaches a predetermined time limit; if the determination result is yes, determining, in the current image, corresponding to the predetermined At least one region of interest of the term as the target region;
  • An embodiment of the present invention further discloses an image storage device, comprising: a determining unit, configured to determine whether a storage time of a current image reaches a predetermined time limit; and a determining unit, configured to: when the determining result is yes, at the current Determining, in the image, at least one region of interest corresponding to the predetermined period as a target region; generating a unit, according to which the target region image containing the target region is generated, wherein The target area image has a smaller amount of data than the current image; the storage unit is configured to store the target area image;
  • Delete unit used to delete the current image.
  • the main differences and effects of the embodiments of the present invention are as follows:
  • the present invention can reduce the range of the target area by extracting the image containing the region of interest, thereby reducing the storage amount of the image.
  • the target area of the image can be gradually reduced, and the purpose of gradually reducing the storage space occupied by the image while retaining the region of interest is realized as the storage time increases. Further, after the original image data is collected, the identification and determination of the region of interest can be performed, and in the subsequent storage process of the image, the data operation for determining the region of interest can be omitted, and the storage cost can be saved.
  • FIG. 1 is a schematic flow chart of an image storage method in a first embodiment of the present invention
  • FIG. 2 is a flow chart showing an image storage method at a vehicle bayonet according to a third embodiment of the present invention
  • FIG. 3 is a schematic diagram of a target area image generated by an image storage method at a vehicle bayonet according to a third embodiment of the present invention
  • FIG. 6 is a schematic structural diagram of an image storage apparatus according to a fourth embodiment of the present invention
  • 7 is a schematic structural diagram of an image storage device according to a fifth embodiment of the present invention
  • FIG. 8 is a schematic structural diagram of an image storage device according to a sixth embodiment of the present invention.
  • Fig. 1 is a flow chart showing the image storage method. Specifically, as shown in Fig. 1, the image storage method includes the following steps: In step 101, it is judged whether or not the storage time of the current image has reached a predetermined deadline. If the result of the determination is yes, proceed to step 102, and if the result of the determination is no, return to the step.
  • the method before step 101, further includes the steps of: forming an original image by using the image of the camera to collect the image; and determining the region of interest in the original image by image recognition and detection technology, and Sense of representation The information of the area of interest is stored in the original image.
  • the identification and determination of the region of interest may be performed immediately after the collection of the original image data, and then the information thereof may be stored in the original image, or may be performed while the image is being stored. .
  • the identification and determination of the region of interest can be performed, and the data operation for determining the region of interest can be omitted in the process of storing the image in the future, thereby saving storage cost.
  • step 102 at least one region of interest corresponding to the predetermined period is determined as a target region in the current image.
  • At least one predetermined period is set, each predetermined period corresponds to a different target area, and the longer the predetermined period, the smaller the target area.
  • the target area of the image can be gradually reduced, and the purpose of gradually reducing the storage space occupied by the image while retaining the region of interest as the storage time increases is realized.
  • one embodiment specifically applied to the bayonet of the vehicle includes two predetermined deadlines, wherein
  • the region of interest corresponding to the first predetermined period is the entire body region.
  • the region of interest corresponding to the second predetermined period of time is the front of the vehicle body region.
  • the second predetermined period is longer than the first predetermined period.
  • the region of interest may also be other information useful in the image for practical applications, such as head information and body information in the face bayonet, or vehicle contours in the vehicle bayonet, body color, vehicle roof, and passengers.
  • the predetermined time limit may be divided according to the needs of the actual application.
  • the region of interest corresponding to the predetermined period is not limited to the division criteria of the embodiment in the present embodiment.
  • a target area image including the target area is generated according to the target area, wherein the data amount of the target area image is smaller than the data amount of the current image.
  • the step 103 further includes the following sub-steps: generating a target area image including only the target area. It can be understood that in other embodiments of the present invention, the storage space of the image in the storage medium can also be reduced by reducing the resolution of the non-target area.
  • the image of the target area is stored.
  • the current image is deleted.
  • a second embodiment of the present invention relates to an image storage method.
  • the present embodiment is an improvement on the basis of the first embodiment.
  • step 103 the image of the target area including the target area is generated according to the target area, and the following sub-steps are further included: The resolution of the target area in the current image; The resolution of the current image that does not contain the portion of the target area is reduced.
  • a third embodiment of the present invention relates to an image storage method at a vehicle bayonet. 2 is a flow chart showing an image storage method at the bayonet of the vehicle.
  • the image identification and detection technologies such as license plate recognition, front detection and window detection are used to determine the key information areas of the vehicle of interest to the criminal investigation, and the images of different regions of interest are hierarchically stored according to different storage periods.
  • the most critical information can be stored for a longer period of time in a limited storage space, and the storage problem of the massive security bayonet image is alleviated. For example, after the original image captured by the security card slot is stored for 3 months, all the information is lost due to the limitation of the storage space, and all information is lost.
  • the image storage method at the vehicle bayonet includes the following steps: In step 201, the capture machine acquires image data and provides the corresponding license plate recognition, front detection, and window detection unit (ie, detection). The unit of the region of interest) as the source data for processing.
  • step 202 the image data acquired by the capture machine is subjected to license plate recognition, and information related to the license plate recognition result, the license plate position, the license plate type, and the like are obtained.
  • step 203 the image data acquired by the capture machine is subjected to front detection to obtain a window position. Set the information.
  • step 204 window image detection is performed on the image data acquired by the capture machine, and information such as the window position is obtained.
  • step 205 based on the acquired information on the license plate, the front, the window, and the like, the related information such as the vehicle area position and the front window position is obtained as a basis for extracting the image of the region of interest.
  • the frame 1 is the front area
  • the frame 2 in the frame 1 is the license plate area
  • the frame 3 is the window area (in order to avoid unnecessary disputes, in the figure)
  • the face and license plate were mosaiced).
  • This information is also available when processing JPEG images on the backend server.
  • step 206 the captured JPEG image is saved, and the acquired information on the vehicle area position, the front window position, and the like are hidden in the JPEG image and stored together with the JPEG image.
  • step 207 it is determined whether the storage time of the image reaches a deadline for grading only the image of the vehicle area (ie, the target area corresponding to the deadline) (ie, the first predetermined period), and if yes, proceeds to step 208, and if not, Continue to judge.
  • step 208 after the image is stored to a certain period (ie, the first predetermined period is reached), the original image is processed, and the vehicle area image is extracted according to information such as the location of the vehicle region hidden in the JPEG image, and only the vehicle area image is saved. Delete the original image to save storage space.
  • An example of the extracted vehicle area image is shown in Figure 4 (to avoid unnecessary disputes, the face and license plate in the picture are mosaic).
  • step 209 when the image is stored to a longer period, it is determined whether the storage time of the image reaches a deadline for grading only the image of the front window area (ie, reaching the second predetermined period), and if yes, proceeding to step 210, if no , then continue to judge.
  • step 210 the region of interest is further reduced, and according to information such as the position of the front window hidden in the JPEG image, the image of the front window area is extracted, and only the image of the front window area is saved, and the image is deleted. In addition to the vehicle area image, the storage space is further saved.
  • the extracted image of the front window area is shown in Figure 5 (to avoid unnecessary disputes, the face and license plate in the picture are mosaic).
  • the present embodiment is not limited to the vehicle bayonet, and is equally applicable to a face mount, the head area corresponding to the front window area, and the entire human body area corresponding to the entire vehicle area.
  • the present embodiment proposes a scheme for grading storage of a large number of security bayonet images.
  • the scheme determines vehicle and vehicle interested in criminal investigation through image recognition and detection technologies such as license plate recognition, vehicle head detection, window detection, face detection and human body detection.
  • image recognition and detection technologies such as license plate recognition, vehicle head detection, window detection, face detection and human body detection.
  • images of different regions of interest are hierarchically stored according to different storage periods.
  • an image region containing key information can be extracted, and the range of the region of interest is gradually reduced, and the amount of storage is reduced, so that the most critical information can be stored for a longer period in a limited storage space, and the image of the large number of security bayonet images is alleviated.
  • This embodiment is a specific application of the first embodiment at the vehicle bayonet.
  • the related technical details mentioned in the first embodiment are still effective in the present embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related art details mentioned in the present embodiment can also be applied to the first embodiment.
  • the method embodiments of the present invention can all be implemented in software, hardware, firmware, and the like.
  • the instruction code can be stored in any type of computer-accessible memory (eg, permanent or modifiable, volatile or non-volatile, solid state Or non-solid, fixed or replaceable media, etc.).
  • the memory may be, for example, Programmable Array Logic ("PAL"), Random Access Memory (RAM), Programmable Read Only Memory (PROM). " ) , Read-Only Memory ( " ROM " ) , Electrically Erasable Programmable ROM ( EEPROM ) , Disk , CD , Digital Versatile Disc , referred to as "DVD”) and so on.
  • PAL Programmable Array Logic
  • RAM Random Access Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • EEPROM Electrically Erasable Programmable ROM
  • Disk Disk
  • CD Digital Versatile Disc
  • DVD Digital Versatile Disc
  • At least one predetermined period is included, each predetermined period corresponds to a different target area, and the longer the predetermined period, the smaller the target area.
  • a determining unit configured to determine, in the current image, at least one region of interest corresponding to the predetermined period as the target region when the determination result is YES.
  • the generating unit generates a target area image including the target area according to the target area, wherein the target area image has a smaller data amount than the current image.
  • a storage unit configured to store the image of the target area.
  • Delete unit used to delete the current image.
  • the apparatus further includes the following unit: a collecting unit, configured to form an original image data by using an image of the camera to collect the points;
  • the region of interest detecting unit is configured to determine a region of interest in the original image by image recognition and detection techniques, and store information representing the region of interest in the original image.
  • the identification and determination of the region of interest may be performed immediately after the collection of the original image data, and then the information thereof is stored in the original image, and the image may also be stored.
  • the identification and determination of the region of interest can be performed, and the data operation for determining the region of interest can be omitted in the process of storing the image in the future, thereby saving storage cost.
  • the apparatus sets two predetermined deadlines.
  • the generating unit of the device includes the following subunit: a first generating subunit, configured to generate and include when the storage time reaches a first predetermined time limit The target area image of the region of interest corresponding to the first predetermined period.
  • the region of interest corresponding to the first predetermined period of time is the entire body region.
  • a second generation subunit configured to generate a target area image including the region of interest corresponding to the second predetermined period when the storage time reaches the second predetermined period.
  • the region of interest corresponding to the second predetermined period of time is the front of the vehicle body region.
  • the third predetermined period of the device may be set, and if the storage time of the image reaches the third predetermined period, the image is deleted.
  • a fifth embodiment of the present invention relates to an image storage device at a vehicle bayonet.
  • Fig. 7 is a schematic structural view of the image storage device.
  • Vehicle related information mainly includes license plate, front (light, heat sink, logo, etc.), body color, vehicle outline, window, vehicle top, etc.
  • the information about the passengers mainly includes the appearance of the cab personnel and the clothes worn.
  • the front and window areas are the most important areas.
  • This area mainly contains the most critical information of criminal investigation: license plate, front, body color, window, appearance of cab personnel and clothes worn;
  • this area contains key information: vehicle outline, vehicle top.
  • the images are divided into three levels depending on the range of the stored image regions.
  • the first-level image captures the original image of the security bayonet, contains the most abundant information, and includes the most important front, window areas and sub-critical car body areas, and also includes the background area outside the vehicle; Contains some redundant background areas that require the most storage space.
  • Secondary image is included The area of the entire vehicle, that is, the most important front, window area and sub-critical car body area, reduces the redundant background area, and the storage of the first-level image greatly reduces the storage space.
  • the third-level image contains only the most important front and window areas mentioned above, and the second-level image retains the most important key information, further reducing the storage area and reducing the storage space.
  • the first-level image area storage period is T1 (for example, 6 months)
  • the second-level image area storage period is T2 (for example, 9 months)
  • the third-level image storage period is T3 (for example, 12 months).
  • the region of the secondary image is extracted from the first-level image, the secondary image is generated, the secondary image is stored, the primary image is deleted, and the time is saved (T2-T1); the secondary image is saved to After the period, the area of the third-level image is extracted from the secondary image, a three-level image is generated, the third-level image is stored, the secondary image is deleted, and the third-level image (T3-T2) time is continuously saved. Therefore, due to the limitation of the storage space, the first-level image is deleted after the T1 time is saved. With the technical solution of the present invention, the image storage space can be greatly reduced and the key information can be saved. Specifically, as shown in FIG.
  • the image storage device mainly includes five modules, which are an image collection unit, a region of interest detection unit, a first-level image generation and storage unit, a secondary image generation and storage unit, and The three-level image generation and storage unit, wherein the region of interest detection unit is further divided into a license plate recognition processing unit, a front detection unit, and a window detection unit.
  • the image collection unit is configured to form an original image data by collecting images of the card point points through the CCD of the camera.
  • the region of interest detection unit is used to identify key information areas such as license plates, fronts, and windows through image recognition and target detection techniques.
  • the region of interest detecting unit mainly includes a license plate recognition processing unit, a head detecting unit, and a window detecting unit.
  • the license plate recognition processing unit is configured to process and recognize the received image data, and identify information such as the license plate position of the vehicle in the image.
  • the front detection unit is configured to receive image data collected by the image collection unit, and receive the image The head is detected like the data to determine the front area.
  • the window detecting unit is configured to receive image data collected by the image collecting unit, perform window detection on the received image data, and determine a window area.
  • the primary image generation and storage unit is configured to generate image data obtained by the image collection unit and store it in an image server, and store information representing the detected region of interest into the primary image.
  • the secondary image generation and storage unit is configured to extract image data of the secondary image region from the primary image according to the region of interest information acquired by the detection unit, generate a secondary image and store the image in the image server, and The information of the area is stored in the secondary image. At the same time, the first level image is deleted.
  • the third-level image generation and storage unit is configured to extract image data of the three-level image region from the secondary image according to the region of interest information acquired by the detecting unit, generate a three-level image and store it in the image server, and represent the region of interest.
  • the information is stored in a three-level image.
  • the secondary image is deleted.
  • the image collection unit collects the image and sends it to the region of interest detection unit.
  • the region of interest detection unit receives the image data collected by the image collection unit, identifies and detects the license plate, the front, and the window, and determines the area in the image of the license plate, the front, and the window.
  • the order of the license plate recognition processing unit, the front detection unit, and the window detection unit in the area of interest detection unit is not required, and may be performed in any order.
  • the front detecting unit in the region of interest detecting unit can be selectively used. If the front detecting unit is not used after detecting the license plate, since the license plate is in the front region, if the license plate is not detected, the front detecting unit must be used to determine the front region. .
  • the region of interest detection unit extracts all regions of interest at once, without the need for hierarchical extraction.
  • the first-level image generation and storage unit generates an image from the image data obtained by the collection unit, and the information of the region of interest obtained by the region of interest detection unit is also hidden in the generated image.
  • the first level image is stored, and the first level storage of the image is completed.
  • the secondary image generation and storage unit generates an image by extracting image data of the secondary image region from the primary image according to the hidden region of interest information in the primary image, and extracts the hidden region of interest information in the primary image. , also hidden into the generated image, storing the secondary image, completing the secondary storage of the image, and deleting the primary image.
  • the three-level image generation and storage unit generates an image by extracting image data of the three-level image region from the secondary image according to the information of the region of interest hidden in the secondary image, storing the three-level image, completing the three-level storage of the image, and deleting the second Level image.
  • a sixth embodiment of the invention relates to an image storage device at a vehicle bayonet.
  • Fig. 8 is a schematic structural view of the image storage device. The present embodiment is improved on the fifth embodiment, and is different from the fifth embodiment in that the detection of the region of interest detecting unit is performed based on the stored first-level image, and is not real-time processing, and the specific processing is no longer performed. Retelling.
  • each unit mentioned in each device implementation manner of the present invention is a logical unit. Physically, one logical unit may be a physical unit, or may be a part of a physical unit, or may have multiple physical parts. The combined implementation of the elements, the physical implementation of these logical units themselves is not the most important, the combination of the functions implemented by these logical units is the key to solving the technical problems raised by the present invention.

Abstract

本发明涉及数据存储领域,公开了一种图像存储方法及其装置。本发明中,包括以下步骤:判断当前图像存储时间是否达到预定期限;如果判断结果为是,则根据该预定期限,在当前图像中确定包含至少一个感兴趣区域的目标区域;根据该目标区域生成包括该目标区域的目标区域图像,其中,该目标区域图像的数据量比当前图像的数据量少;存储目标区域图像;删除当前图像。随着存储时间的增长,通过设置不同的期限,逐步去掉图像中冗余的信息,可以减小图像占用的存储空间,同时保留各种实际应用中所需的感兴趣区域的信息,实现在有限的存储空间上尽可能长时间存储关键信息的目的。

Description

说 明 书
图像存储方法及其装置 技术领域 本发明涉及数据存储领域, 特别涉及一种图像存储技术。
背景技术 治安卡口是指依托道路上特定场所, 如收费站、 交通或治安检查站等卡 口点, 对所有通过该卡口点的机动车辆进行拍摄、 记录与处理的一种道路交 通现场监测系统。 当前, 高清卡口系统在治安、 刑侦、 打击和预防犯罪、 处 置突发性事件等应用中发挥重要作用, 特别在刑侦应用中需要根据存储的卡 口图像寻找破案线索, 这种方法已是刑侦破案的重要技术手段。 刑侦破案可能需要查看三个月或更长时间以前的历史图像,这就要求治安 卡口图像保存较长时间, 而现在高清卡口建设越来越多, 大数量卡口图像保 存就需要非常大的存储空间。 高清治安卡口系统目前面临的问题就是: 大数 量卡口图像长期存储的问题。 中国专利 CN201210568309为本发明最接近的技术方案, 其公开了一种 降低存储空间的方法。 该发明提出一种降低图像存储空间的方法, 解决现有 技术中监控系统需要大存储空间来存储图像信号的问题。 其技术方案为: 通过摄像头获取图像信号, 将图像信号进行緩冲, 输出 数据信号, 对数据信号进行压缩, 提取 GIF ( Graphics Interchange Format, 图像互换格式) 或者 JPG ( Joint Photographic Experts Group , 联合图像专家 小组)格式的图像数据, 将最新的图像数据与寄存器中上一次更新的图像数 据相比较, 如果比较结果超出了范围, 提取 JPG格式的图像数据, 否则提取 GIF格式的图像数据。 其技术方案的本质就是: 在获取的图像没有差异的情况下, 将低像素的 图像存储在存储器中, 降低了存储空间。 这种通过降低图像分辨率的方式来 减少存储空间的方法, 尽管减少了存储空间, 但也降低了关键信息的可辨识 度, 这种关键信息的丟失, 会严重影响治安卡口图像在刑侦中的应用。
发明内容
本发明的目的在于提供一种图像存储方法及其装置, 通过设置不同的期 限, 逐步去掉图像中冗余或者不重要的信息, 减小图像占用的存储空间, 同 时, 保留各种实际应用中所需的感兴趣区域的信息。
为解决上述技术问题, 本发明的实施方式公开了一种图像存储方法, 包 括以下步骤: 判断当前图像的存储时间是否达到预定期限; 如果判断结果为是, 则在当前图像中确定对应于该预定期限的至少一个 感兴趣区域作为目标区域;
根据该目标区域生成包括该目标区域的目标区域图像, 其中, 该目标区 域图像的数据量比当前图像的数据量少;
存储该目标区域图像; 删除当前图像。 本发明的实施方式还公开了一种图像存储装置, 包括以下单元: 判断单元, 用于判断当前图像的存储时间是否达到预定期限; 确定单元, 用于在所述判断结果为是时, 在当前图像中确定对应于该预 定期限的至少一个感兴趣区域作为目标区域; 生成单元,根据该目标区域生成包含该目标区域的目标区域图像,其中, 该目标区域图像的数据量比当前图像的数据量少; 存储单元, 用于存储该目标区域图像;
删除单元, 用于删除当前图像。 本发明实施方式与现有技术相比, 主要区别及其效果在于:
随着图像存储时间的增长, 通过设置不同的期限, 逐步去掉图像中冗余 或者不重要的信息, 能够减小图像占用的存储空间, 同时, 保留各种实际应 用中所需的感兴趣区域的信息。在各种治安卡口,如车辆卡口和人脸卡口中, 使用本发明, 能够通过提取包含感兴趣区域的图像, 随时间逐步减少目标区 域的范围, 从而减少图像的存储量, 达到在有限的存储空间上尽可能长时间 存储关键信息的目的, 进而有效解决大数量图像存储空间不足的问题, 降低 存储成本。
进一步地, 通过设置多个预定期限, 可以逐步减小图像的目标区域, 实 现随着存储时间的增长, 在保留感兴趣区域的同时逐步减少图像所占的存储 空间的目的。 进一步地, 在原始图像数据釆集后进行感兴趣区域的识别和确定, 可以 在以后图像的存储处理过程中, 省去确定感兴趣区域的数据运算, 节约存储 成本。
进一步地, 通过图像中降低非目标区域的分辨率, 不仅可以减小该图像 在存储介质中的存储空间, 而且, 能够在保留感兴趣区域的同时保留感兴趣 区域的相关背景, 最大限度的保留图像显示的信息。
附图说明
图 1是本发明第一实施方式中一种图像存储方法的流程示意图; 图 2是本发明第三实施方式中一种车辆卡口处的图像存储方法的流程示 意图;
图 3至图 5是本发明第三实施方式中一种车辆卡口处的图像存储方法生 成的目标区域图像的示意图; 图 6是本发明第四实施方式中一种图像存储装置的结构示意图; 图 7是本发明第五实施方式中一种图像存储装置的结构示意图; 图 8是本发明第六实施方式中一种图像存储装置的结构示意图。
具体实施方式
在以下的叙述中, 为了使读者更好地理解本申请而提出了许多技术细 节。 但是, 本领域的普通技术人员可以理解, 即使没有这些技术细节和基于 以下各实施方式的种种变化和修改, 也可以实现本申请各权利要求所要求保 护的技术方案。 为使本发明的目的、 技术方案和优点更加清楚, 下面将结合附图对本发 明的实施方式作进一步地详细描述。 本发明第一实施方式涉及一种图像存储方法。 图 1是该图像存储方法的 流程示意图。 具体地说, 如图 1所示, 该图像存储方法包括以下步骤: 在步骤 101 中, 判断当前图像的存储时间是否达到预定期限。 如果判断结果为是, 则进入步骤 102, 如果判断结果为否, 则回到步骤
101。 本实施方式中, 在步骤 101之前, 还包括以下步骤: 通过摄像机的釆集卡口点釆集的图像, 形成原始图像; 通过图像识别和检测技术, 确定原始图像中的感兴趣区域, 并将代表感 兴趣区域的信息存储在原始图像中。
可以理解, 在本发明的其他实施方式中, 感兴趣区域的识别和确定可以 在原始图像数据釆集后立即进行, 然后将其信息存储到原始图像中, 也可以 在对图像进行存储处理时进行。
在原始图像数据釆集后进行感兴趣区域的识别和确定, 可以在以后图像 的存储处理过程中, 省去确定感兴趣区域的数据运算, 节约存储成本。
在步骤 102中, 在当前图像中确定对应于该预定期限的至少一个感兴趣 区域作为目标区域。
在本实施方式中, 设定至少一个预定期限, 每个预定期限对应不同的目 标区域, 且预定期限越长, 目标区域越小。
通过设置多个预定期限, 可以逐步减小图像的目标区域, 实现随着存储 时间的增长,在保留感兴趣区域的同时逐步减少图像所占的存储空间的目的。
优选地, 在本实施方式中一个具体应用于车辆卡口处的实施例中, 包含 两个预定期限, 其中,
对应第一预定期限的感兴趣区域为整个车身区域。
对应第二预定期限的感兴趣区域为该车身区域中的车头。 其中, 该第二预定期限比该第一预定期限长。 可以理解, 上述整个车身区域包含车辆的车头、 车窗和车窗后的车身区 域, 而上述车身区域的车头包含车辆的车窗、 车牌。 感兴趣区域也可以是图像中对实际应用有用的其他信息, 比如人脸卡口 中人头信息和人体信息, 或车辆卡口中的车辆轮廓、 车身颜色、 车辆顶部、 以及司乘人员等信息。 在本发明的其他实施方式中, 可以根据实际应用的需要, 划分预定期限 和预定期限相对应的感兴趣区域,不限于本实施方式中该实施例的划分标准。 此后进入步骤 1 03, 根据该目标区域生成包括该目标区域的目标区域图 像, 其中, 该目标区域图像的数据量比当前图像的数据量少。 在本实施方式中, 本步骤 1 03还包括以下子步骤: 生成仅包含上述目标区域的目标区域图像。 可以理解, 在本发明的其他实施方式中, 也可以通过降低非目标区域的 分辨率降低图像在存储介质中的存储空间。 此后进入步骤 104, 存储该目标区域图像。 此后进入步骤 105, 删除当前图像。 此外, 可以理解, 在当前图像的存储时间未到达预定期限时, 不进行任 何处理, 保持当前图像不变。 随着图像存储时间的增长, 通过设置不同的期限, 逐步去掉图像中冗余 或者不重要的信息, 能够减小图像占用的存储空间, 同时, 保留各种实际应 用中所需的感兴趣区域的信息。在各种治安卡口,如车辆卡口和人脸卡口中, 使用本发明, 能够通过提取包含感兴趣区域的图像, 随时间逐步减少目标区 域的范围, 从而减少图像的存储量, 达到在有限的存储空间上尽可能长时间 存储关键信息的目的, 进而有效解决大数量图像存储空间不足的问题, 降低 存储成本。 本发明第二实施方式涉及一种图像存储方法。 本实施方式是在第一实施 方式的基础上做了改进, 主要改进之处在于: 在步骤 1 03中,即根据该目标区域生成包括该目标区域的目标区域图像, 还包含以下子步骤: 保持当前图像中目标区域的分辨率; 降低该当前图像中不包含该目标区域部分的分辨率。
通过降低图像中非目标区域的分辨率, 不仅可以减小该图像在存储介质 中的存储空间, 而且, 能够在保留感兴趣区域的同时保留感兴趣区域的相关 背景, 最大限度的保留图像显示的信息。 第一实施方式中提到的相关技术细节在本实施方式中依然有效, 为了减 少重复, 这里不再赘述。 本发明第三实施方式涉及一种车辆卡口处的图像存储方法。 图 2是该车 辆卡口处的图像存储方法的流程示意图。 本实施方式中, 通过车牌识别、 车头检测及车窗检测等图像识别和检测 技术确定刑侦感兴趣的车辆关键信息区域, 按照不同存储周期分级存储不同 感兴趣区域图像。 这样可以通过逐步减少感兴趣区域的范围, 减少存储量, 从而达到在有限存储空间上, 最关键信息可以存储周期更长, 緩解海量治安 卡口图像的存储问题。 例如, 治安卡口抓拍的原始图像存储 3个月后, 由于存储空间的限制就 全部删除, 所有信息丟失。 本实施方式可在存储 3个月后, 将车辆区域提取 出来, 仅保存车辆区域图像, 6 个月后再提取车窗车头部分, 仅存储车窗和 车头区域图像, 大大减少了存储空间。 逐步去掉冗余或不重要的信息, 逐步 减小存储图像区域, 实现分级存储, 使得最关键的车辆信息存储周期更久。 具体地, 如图 2所示, 该车辆卡口处的图像存储方法包括以下步骤: 在步骤 201 中, 抓拍机获取图像数据, 提供给相应的车牌识别、 车头检 测和车窗检测单元 (即检测感兴趣区域的单元)作为处理的源数据。 在步骤 202中, 对抓拍机获取的图像数据, 进行车牌识别, 获得车牌识 别结果、 车牌位置、 车牌类型等相关信息。 在步骤 203中, 对抓拍机获取的图像数据, 进行车头检测, 获得车窗位 置等信息。
在步骤 204中, 对抓拍机获取的图像数据, 进行车窗检测, 获得车窗位 置等信息。
在步骤 205中, 根据获取的车牌、 车头、 车窗等相关信息进行分析, 得 到车辆区域位置和车头车窗位置等相关信息, 作为提取感兴趣区域图像的基 础。 如图 3所示, 在包含获取的感兴趣区域的图像中, 框 1为车头区域, 框 1 内的框 2为车牌区域, 框 3为车窗区域(为了避免引起不必要的纠纷, 图 中的人脸和车牌打了马赛克) 。 这些信息也可在后端服务器上对 JPEG图像进行处理时获取。 在步骤 206中, 保存抓拍到的 JPEG图像, 并将获取的车辆区域位置、 车头车窗位置等相关信息隐藏于 JPEG图像中,与 JPEG图像一起存储起来。 在步骤 207中, 判断图像的存储时间是否达到分级仅保存车辆区域(即 对应于该期限的目标区域) 图像的期限(即第一预定期限) , 如果是, 则进 入步骤 208, 如果否, 则继续判断。 在步骤 208中, 在图像存储到一定周期后 (即达到第一预定期限)对原 图像进行处理, 根据隐藏于 JPEG图像中的车辆区域位置等信息, 提取车辆 区域图像, 仅保存车辆区域图像, 删除原图像, 节省存储空间。 提取的车辆 区域图像示例如图 4所示 (为了避免引起不必要的纠纷, 图中的人脸和车牌 打了马赛克) 。
在步骤 209中, 当图像存储到更长周期时, 判断图像的存储时间是否达 到分级仅保存车头车窗区域图像的期限(即达到第二预定期限) , 如果是, 则进入步骤 210, 如果否, 则继续判断。 在步骤 210中, 感兴趣区域进一步减小, 根据隐藏于 JPEG图像中的车 头车窗位置等信息, 提取车头车窗区域图像, 仅保存车头车窗区域图像, 删 除车辆区域图像, 进一步节省了存储空间。 提取的车头车窗区域图像如图 5 所示 (为了避免引起不必要的纠纷, 图中的人脸和车牌打了马赛克) 。 本实施方式不局限于车辆卡口, 同样适用于人脸卡口, 人头区域相当于 上述车头车窗区域, 整个人体区域相当于上述整个车辆区域。 本实施方式提出了一种大数量治安卡口图像分级存储的方案, 该方案通 过车牌识别、 车头检测、 车窗检测、 人脸检测及人体检测等图像识别和检测 技术确定刑侦感兴趣的车辆和人体关键信息区域, 按照不同存储周期分级存 储不同感兴趣区域图像。 本实施方式可提取包含关键信息的图像区域, 并逐 步减少感兴趣区域的范围, 减少存储量, 从而达到在有限存储空间上, 最关 键信息可以存储周期更长, 緩解大数量治安卡口图像的存储问题。 本实施方式是第一实施方式在车辆卡口处的具体应用。 第一实施方式中 提到的相关技术细节在本实施方式中依然有效, 为了减少重复, 这里不再赘 述。相应地,本实施方式中提到的相关技术细节也可应用在第一实施方式中。 本发明的各方法实施方式均可以以软件、 硬件、 固件等方式实现。 不管 本发明是以软件、 硬件、 还是固件方式实现, 指令代码都可以存储在任何类 型的计算机可访问的存储器中 (例如永久的或者可修改的, 易失性的或者非 易失性的, 固态的或者非固态的, 固定的或者可更换的介质等等) 。 同样, 存储器可以例如是可编程阵列逻辑 ( Programmable Array Logic , 简称 "PAL" ) 、 随机存取存储器(Random Access Memory, 简称 " RAM" ) 、 可编程只读存储器( Programmable Read Only Memory, 简称 "PROM" ) 、 只读存储器( Read-Only Memory, 简称 " ROM" ) 、 电可擦除可编程只读 存储器( Electrically Erasable Programmable ROM , 简称 "EEPROM" ) 、 磁盘、 光盘、 数字通用光盘 ( Digital Versatile Disc, 简称 "DVD" ) 等等。 本发明第四实施方式涉及一种图像存储装置。 图 6是该图像存储装置的 结构示意图。 具体地说, 如图 6所示, 该图像存储装置包括以下单元: 判断单元, 用于判断当前图像的存储时间是否达到预定期限。
在本实施方式中, 包含至少一个预定期限, 每个预定期限对应不同的目 标区域, 且预定期限越长, 目标区域越小。 确定单元, 用于在判断结果为是时, 在当前图像中确定对应于该预定期 限的至少一个感兴趣区域作为目标区域。
生成单元,根据该目标区域生成包含该目标区域的目标区域图像,其中, 该目标区域图像的数据量比当前图像的数据量少。 存储单元, 用于存储该目标区域图像。
删除单元, 用于删除当前图像。
在本实施方式中, 该装置还包括以下单元: 釆集单元, 用于通过摄像机的釆集卡口点釆集的图像, 形成原始图像数 据;
感兴趣区域检测单元, 用于通过图像识别和检测技术, 确定原始图像中 的感兴趣区域, 并将代表该感兴趣区域的信息存储在原始图像中。
此外, 可以理解, 在本发明的其他实施方式中, 感兴趣区域的识别和确 定可以在原始图像数据釆集后立即进行,然后将其信息存储到原始图像的中, 也可以在对图像进行存储处理时进行。 在原始图像数据釆集后进行感兴趣区域的识别和确定, 可以在以后图像 的存储处理过程中, 省去确定感兴趣区域的数据运算, 节约存储成本。
优选地, 在本实施方式应用于车辆卡口处的一个具体实施例中, 该装置 设定两个预定期限。 且该装置的生成单元包括以下子单元: 第一生成子单元, 用于在存储时间达到第一预定期限时, 生成包含与该 第一预定期限对应的感兴趣区域的目标区域图像。 与第一预定期限对应的感 兴趣区域为整个车身区域。 第二生成子单元, 用于在存储时间达到第二预定期限时, 生成包含与该 第二预定期限对应的感兴趣区域的目标区域图像。 与第二预定期限对应的感 兴趣区域为该车身区域中的车头。
其中, 第二预定期限比第一预定期限长。 在本实施例中, 可以设定装置的第三预定期限, 如果图像的存储时间达 到该第三预定期限, 则删除图像。
第一实施方式互相配合实施。 第一实施方式中提到的相关技术细节在本实施 方式中依然有效, 为了减少重复, 这里不再赘述。 相应地, 本实施方式中提 到的相关技术细节也可应用在第一实施方式中。 本发明第五实施方式涉及一种车辆卡口处的图像存储装置。 图 7是该图 像存储装置的结构示意图。 从刑侦角度看, 治安卡口图像中关键信息主要包含车辆和司乘人员的相 关信息。 车辆相关信息主要包括车牌、 车头 (车灯、 散热片、 车标等) 、 车 身颜色、 车辆轮廓、 车窗、 车辆顶部等; 司乘人员相关信息主要包括驾驶室 人员的外貌和所穿衣服。 从信息的重要程度看, 车头和车窗区域为最重要的 区域, 此区域主要包含刑侦最关键信息: 车牌、 车头、 车身颜色、 车窗、 驾 驶室人员的外貌和所穿衣服; 次重要区域为车窗以后的车身区域, 此区域包 含比较关键信息: 车辆轮廓、 车辆顶部。 在本实施方式中, 根据存储图像区域范围不同, 将图像分为三级。 一级 图像为治安卡口抓拍的原始图像, 包含信息最丰富, 同时包含了上述的最重 要车头、 车窗区域和次重要的车窗以后的车身区域, 还包含了车辆以外的背 景区域; 其包含一些冗余的背景区域, 需要存储空间最大。 二级图像为包含 整个车辆的区域, 即上述的最重要车头、 车窗区域和次重要的车窗以后的车 身区域, 其减少了冗余的背景区域, 相对一级图像的存储, 大大减少了存储 空间。 三级图像为仅包含上述的最重要车头、 车窗区域, 相对二级图像, 保 留了最重要的关键信息, 进一步减小存储区域, 减少了存储空间。
基于上述理念, 在存储介质 (例如, 硬盘) 上可按照不同存储周期分级 存储。 例如, 上述一级图像区域存储周期为 T1 (例如, 6个月 ) , 上述二级 图像区域存储周期为 T2 (例如, 9个月 ) , 三级图像存储周期为 T3 (例如, 12个月 ) 。 上述一级图像保存 T1 时间后, 从一级图像中提取二级图像的区 域, 生成二级图像, 存储二级图像, 删除一级图像, 继续保存 (T2-T1 ) 时 间; 二级图像保存到期后, 从二级图像中提取三级图像的区域, 生成三级图 像, 存储三级图像, 删除二级图像, 继续保存三级图像 (T3-T2 ) 时间。 这 样原来由于存储空间限制, 一级图像保存 T1 时间后, 就会全部删除, 通过 本发明的技术方案, 可以大大减少图像存储空间并使关键信息继续保存。 具体地说, 如图 7所示, 该图像存储装置主要包括 5个模块, 分别是图 像釆集单元、 感兴趣区域检测单元、 一级图像生成与存储单元、 二级图像生 成与存储单元、 以及三级图像生成与存储单元, 其中感兴趣区域检测单元又 分为车牌识别处理单元、 车头检测单元、 车窗检测单元。 图像釆集单元用于通过摄像机的 CCD釆集卡口点的图像, 形成原始的 图像数据。 感兴趣区域检测单元用于通过图像识别和目标检测技术确车牌、 车头、 车窗等关键信息区域。 并且, 感兴趣区域检测单元主要包括车牌识别处理单 元、 车头检测单元、 车窗检测单元。
其中, 车牌识别处理单元用于对接收到的图像数据进行处理和识别, 识 别出图像中车辆的车牌位置等信息。 车头检测单元用于接收图像釆集单元釆集到的图像数据, 对接收到的图 像数据进行车头检测, 确定车头区域。
车窗检测单元用于接收图像釆集单元釆集到的图像数据, 对接收到的图 像数据进行车窗检测, 确定车窗区域。
一级图像生成与存储单元用于将所述图像釆集单元获得的图像数据生 成图像并存储到图像服务器中, 并将代表检测到的感兴趣区域的信息存储到 一级图像中。
二级图像生成与存储单元用于根据检测单元获取的感兴趣区域信息, 从 一级图像中提取二级图像区域的图像数据, 生成二级图像并存储到图像服务 器中, 并并将代表感兴趣区域的信息存储到二级图像中。 同时, 删除一级图 像。
三级图像生成与存储单元用于根据检测单元获取的感兴趣区域信息, 从 二级图像中提取三级图像区域的图像数据, 生成三级图像并存储到图像服务 器中, 并将代表感兴趣区域的信息存储到三级图像中。 同时, 删除二级图像。
该图像存储装置的实时进行方式如图 7所示:
首先, 图像釆集单元釆集图像, 送入感兴趣区域检测单元。 其次,感兴趣区域检测单元接收图像釆集单元釆集的图像数据,对车牌、 车头、 车窗进行识别和检测, 确定车牌、 车头、 车窗所在图像中的区域。 感 兴趣区域检测单元中的车牌识别处理单元、 车头检测单元、 车窗检测单元的 先后顺序没有要求, 可按任意顺序进行。 感兴趣区域检测单元中的车头检测 单元可选择性地使用, 若检测到车牌后可不使用车头检测单元, 因为车牌就 在车头区域内, 若未检测到车牌, 就必须使用车头检测单元确定车头区域。 感兴趣区域检测单元一次性提取所有兴趣区域, 无需分级提取。
然后, 一级图像生成与存储单元将釆集单元获得的图像数据生成图像, 同时感兴趣区域检测单元获得的感兴趣区域的信息也隐藏于生成的图像存 中, 存储一级图像, 完成图像的一级存储。 然后, 二级图像生成与存储单元根据一级图像中隐藏的感兴趣区域信息 从一级图像中提取二级图像区域的图像数据生成图像, 同时将一级图像中隐 藏的感兴趣区域信息提取出来, 也隐藏到生成的图像中, 存储二级图像, 完 成图像的二级存储, 删除一级图像。 最后, 三级图像生成与存储单元根据二级图像中隐藏的感兴趣区域信息 从二级图像中提取三级图像区域的图像数据生成图像, 存储三级图像, 完成 图像的三级存储, 删除二级图像。 本发明第六实施方式涉及一种车辆卡口处的图像存储装置。 图 8是该图 像存储装置的结构示意图。 本实施方式是在第五实施方式上进行了改进, 与第五实施方式的不同在 于, 感兴趣区域检测单元的检测是基于已存储的一级图像进行的, 非实时处 理, 具体处理过程不再重述。 基于已存储的一级图像进行感兴趣区域的检测, 可以在提取二级图像数 据时, 根据当时的需要确定感兴趣区域, 从而, 确定提高感兴趣区域的灵活 性和实时性。 在本发明的其他实施方式, 也可以根据实际需要, 在存储各级图像时实 时进行。 需要说明的是, 本发明各设备实施方式中提到的各单元都是逻辑单元, 在物理上, 一个逻辑单元可以是一个物理单元, 也可以是一个物理单元的一 部分, 还可以以多个物理单元的组合实现, 这些逻辑单元本身的物理实现方 式并不是最重要的, 这些逻辑单元所实现的功能的组合才是解决本发明所提 出的技术问题的关键。 此外, 为了突出本发明的创新部分, 本发明上述各设 备实施方式并没有将与解决本发明所提出的技术问题关系不太密切的单元引 需要说明的是, 在本专利的权利要求和说明书中, 诸如第一和第二等之 类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来, 而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺 序。 而且, 术语 "包括" 、 "包含" 或者其任何其他变体意在涵盖非排他性 的包含, 从而使得包括一系列要素的过程、 方法、 物品或者设备不仅包括那 些要素, 而且还包括没有明确列出的其他要素, 或者是还包括为这种过程、 方法、 物品或者设备所固有的要素。 在没有更多限制的情况下, 由语句 "包 括一个" 限定的要素, 并不排除在包括所述要素的过程、 方法、 物品或者设 备中还存在另外的相同要素。 虽然通过参照本发明的某些优选实施方式, 已经对本发明进行了图示和 描述, 但本领域的普通技术人员应该明白, 可以在形式上和细节上对其作各 种改变, 而不偏离本发明的精神和范围。

Claims

权 利 要 求 书
1. 一种图像存储方法, 其特征在于, 包括以下步骤:
判断当前图像的存储时间是否达到预定期限;
如果判断结果为是, 则在当前图像中确定对应于该预定期限的至少一个 感兴趣区域作为目标区域;
根据该目标区域生成包括该目标区域的目标区域图像, 其中, 该目标区 域图像的数据量比所述当前图像的数据量少;
存储该目标区域图像;
删除当前图像。
2. 根据权利要求 1 所述的图像存储方法, 其特征在于, 设定至少一个 预定期限, 每个预定期限对应不同的目标区域, 且预定期限越长, 所述目标 区域越小。
3. 根据权利要求 2 所述的图像存储方法, 其特征在于, 在所述判断当 前图像的存储时间是否达到预定期限的步骤之前, 还包括以下步骤:
通过摄像机的釆集卡口点釆集的图像, 形成原始图像;
通过图像识别和检测技术, 确定原始图像中的感兴趣区域, 并将代表所 述感兴趣区域的信息存储在原始图像中。
4. 根据权利要求 3 所述的图像存储方法, 其特征在于, 该方法用于车 辆卡口处图像的存储, 且该方法包含两个所述预定期限, 其中,
对应第一预定期限的感兴趣区域为整个车身区域;
对应第二预定期限的感兴趣区域为该车身区域中的车头; 该第二预定期限比该第一预定期限长。
5. 根据权利要求 1 至 4中任一项所述的图像存储方法, 其特征在于, 所述根据该目标区域生成包含所述目标区域的目标区域图像的步骤包括以下 子步骤:
生成仅包含所述目标区域的目标区域图像。
6. 根据权利要求 1至 4中任一项所述的图像存储方法,其特征在于, 所 述根据该目标区域生成包含所述目标区域的目标区域图像的步骤包括以下子 步骤: 保持所述当前图像中所述目标区域的分辨率;
降低该当前图像中不包含该目标区域部分的分辨率。
7. 一种图像存储装置, 其特征在于, 包括以下单元: 判断单元, 用于判断当前图像的存储时间是否达到预定期限;
确定单元, 用于在所述判断结果为是时, 在当前图像中确定对应于该预 定期限的至少一个感兴趣区域作为目标区域; 生成单元,根据该目标区域生成包含该目标区域的目标区域图像,其中, 该目标区域图像的数据量比所述当前图像的数据量少; 存储单元, 用于存储该目标区域图像;
删除单元, 用于删除当前图像。
8. 根据权利要求 7 所述的图像存储装置, 其特征在于, 包含至少一个 预定期限, 每个预定期限对应不同的目标区域, 且预定期限越长, 所述目标 区域越小。
9. 根据权利要求 8 所述的图像存储装置, 其特征在于, 还包括以下单 元: 釆集单元, 用于通过摄像机的釆集卡口点釆集的图像, 形成原始图像数 据; 感兴趣区域检测单元, 用于通过图像识别和检测技术, 确定原始图像中 的感兴趣区域, 并将代表该感兴趣区域的信息存储在原始图像中。
1 0. 根据权利要求 9所述的图像存储装置, 其特征在于, 该装置用于储 存车辆卡口处的图像, 且该装置设定两个所述预定期限, 且所述生成单元包 括以下子单元: 第一生成子单元, 用于在存储时间达到第一预定期限时, 生成包含与该 第一预定期限对应的感兴趣区域的目标区域图像; 第二生成子单元, 用于在存储时间达到第二预定期限时, 生成包含与该 第二预定期限对应的感兴趣区域的目标区域图像; 其中, 所述第二预定期限比所述第一预定期限长。
1 1 . 根据权利要求 1 0所述的图像存储装置, 其特征在于, 所述与第一预定期限对应的感兴趣区域为整个车身区域; 所述与第二预定期限对应的感兴趣区域为该车身区域中的车头。
PCT/CN2014/084294 2013-09-04 2014-08-13 图像存储方法及其装置 WO2015032267A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP14841969.0A EP3043555B1 (en) 2013-09-04 2014-08-13 Image storage method and device thereof
US14/916,893 US9870509B2 (en) 2013-09-04 2014-08-13 Image storage method and apparatus thereof

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310398902.0 2013-09-04
CN201310398902.0A CN104427310B (zh) 2013-09-04 2013-09-04 图像存储方法及其装置

Publications (1)

Publication Number Publication Date
WO2015032267A1 true WO2015032267A1 (zh) 2015-03-12

Family

ID=52627782

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/084294 WO2015032267A1 (zh) 2013-09-04 2014-08-13 图像存储方法及其装置

Country Status (4)

Country Link
US (1) US9870509B2 (zh)
EP (1) EP3043555B1 (zh)
CN (1) CN104427310B (zh)
WO (1) WO2015032267A1 (zh)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105869118A (zh) * 2016-04-19 2016-08-17 北京君和信达科技有限公司 安检服务器、移动检查终端、系统及处理安检图像的方法
CN106060382A (zh) * 2016-05-27 2016-10-26 北京金山安全软件有限公司 一种图像处理方法、装置及电子设备
DE102016213493A1 (de) * 2016-07-22 2018-01-25 Conti Temic Microelectronic Gmbh Kameravorrichtung zur Aufnahme eines Umgebungsbereichs eines eigenen Fahrzeugs sowie Verfahren zur Bereitstellung einer Fahrerassistenzfunktion
US10395335B2 (en) * 2017-03-27 2019-08-27 International Business Machines Corporation Optimal data sampling for image analysis
CN108195855B (zh) * 2017-12-27 2023-11-03 同方威视技术股份有限公司 安全检查系统及其方法
US10713518B2 (en) * 2018-01-29 2020-07-14 Toyota Motor Engineering & Manufacturing North America, Inc. System and method for recording and reporting license number violation
CN111163258B (zh) * 2018-11-07 2023-04-18 中兴通讯股份有限公司 一种图像存储控制方法、设备和存储介质
CN109600565B (zh) * 2018-12-13 2021-02-26 广州路派电子科技有限公司 一种基于泊车系统的录像方法
CN110099251A (zh) * 2019-04-29 2019-08-06 努比亚技术有限公司 监控视频的处理方法、装置以及计算机可读存储介质
CN112770117B (zh) * 2020-12-30 2022-09-06 杭州海康机器人技术有限公司 视频存储方法、处理方法、装置和存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101112099A (zh) * 2005-02-01 2008-01-23 松下电器产业株式会社 监控记录装置
CN101674444A (zh) * 2009-10-15 2010-03-17 杭州华三通信技术有限公司 视频数据的存储方法、系统和编码器
CN101783852A (zh) * 2009-01-21 2010-07-21 兄弟工业株式会社 读取装置和通信系统
CN102801955A (zh) * 2011-08-17 2012-11-28 南京金柏图像技术有限公司 一种基于局部高清的数字视频传输方法
CN103020624A (zh) * 2011-09-23 2013-04-03 杭州海康威视系统技术有限公司 混合车道监控视频智能标记、检索回放方法及其装置
CN103051877A (zh) 2012-12-12 2013-04-17 青岛联盟电子仪器有限公司 一种降低图像存储空间方法
CN103258028A (zh) * 2013-05-08 2013-08-21 林凡 一种基于内容特征的视频分级分区存储系统

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8213041B2 (en) * 2006-09-25 2012-07-03 Kyocera Mita Corporation Image forming apparatus provided with an image filing function for saving files in an organized manner
WO2008075779A1 (en) * 2006-12-18 2008-06-26 Fujifilm Corporation Monitoring system, monitoring method and program
CN101394531A (zh) * 2007-09-17 2009-03-25 映佳科技股份有限公司 低数据量录像影像储存方法
CN102413301A (zh) * 2010-09-21 2012-04-11 富泰华工业(深圳)有限公司 图像存储控制装置及方法、读取、测试系统及方法
CN102164231A (zh) * 2011-01-24 2011-08-24 南京壹进制信息技术有限公司 基于敏感信息的分辨率可动态调节的监控装置及方法
JP5652408B2 (ja) * 2012-01-16 2015-01-14 カシオ計算機株式会社 撮像装置、撮像方法及びプログラム

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101112099A (zh) * 2005-02-01 2008-01-23 松下电器产业株式会社 监控记录装置
CN101783852A (zh) * 2009-01-21 2010-07-21 兄弟工业株式会社 读取装置和通信系统
CN101674444A (zh) * 2009-10-15 2010-03-17 杭州华三通信技术有限公司 视频数据的存储方法、系统和编码器
CN102801955A (zh) * 2011-08-17 2012-11-28 南京金柏图像技术有限公司 一种基于局部高清的数字视频传输方法
CN103020624A (zh) * 2011-09-23 2013-04-03 杭州海康威视系统技术有限公司 混合车道监控视频智能标记、检索回放方法及其装置
CN103051877A (zh) 2012-12-12 2013-04-17 青岛联盟电子仪器有限公司 一种降低图像存储空间方法
CN103258028A (zh) * 2013-05-08 2013-08-21 林凡 一种基于内容特征的视频分级分区存储系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3043555A4

Also Published As

Publication number Publication date
EP3043555A4 (en) 2016-08-24
EP3043555B1 (en) 2018-09-26
US9870509B2 (en) 2018-01-16
US20160210517A1 (en) 2016-07-21
EP3043555A1 (en) 2016-07-13
CN104427310B (zh) 2017-11-10
CN104427310A (zh) 2015-03-18

Similar Documents

Publication Publication Date Title
WO2015032267A1 (zh) 图像存储方法及其装置
US8953044B2 (en) Multi-resolution video analysis and key feature preserving video reduction strategy for (real-time) vehicle tracking and speed enforcement systems
Wahab et al. Passive video forgery detection techniques: A survey
US20120300073A1 (en) Application server and method for finding parking space
EP2662827B1 (en) Video analysis
CN105336170A (zh) 交通违法监测方法及其装置
JP2009147911A (ja) 映像データ圧縮前処理方法およびこれを用いた映像データ圧縮方法と映像データ圧縮システム
WO2014193220A2 (en) System and method for multiple license plates identification
CN102905101A (zh) 映像记录装置、映像记录系统以及映像记录方法
CN111402298A (zh) 基于目标检测与轨迹分析的粮库视频数据压缩方法
JP2008005399A (ja) 放置物検出装置及び放置物検出方法
CA3101864A1 (en) Systems and methods for identifying vehicles using wireless device identifiers
JP7255819B2 (ja) 映像ストリームからの物体検出において用いるためのシステム及び方法
JP2006202000A (ja) 車両監視方法及び特定文字パターン認識装置及び車両監視システム
KR102506971B1 (ko) 딥러닝 기반의 후방 촬영을 통한 이륜자동차 번호판 인식 방법 및 시스템
KR102042397B1 (ko) 압축영상에 대한 신택스 기반의 히트맵 생성 방법
CN102881161A (zh) 基于多帧差分与投射阴影去除的运动车辆检测方法及装置
CN103905789A (zh) 动态分析数据的保存、搜索和事后分析的方法及其装置
US11804046B2 (en) Systems and methods for efficient information extraction in traffic monitoring
Sridevi et al. Automatic generation of traffic signal based on traffic volume
TWI617997B (zh) Intelligent object detection assistance system and method
WO2015180252A1 (zh) 一种多目标逆向违规检测方法、装置及系统
Roy et al. Vehicle number plate recognition and parking system
US20230281984A1 (en) Method, system and computer program product for correlating objects of interest based on difference and similarities analysis
Kaur et al. Automatic license plate recognition system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14841969

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 14916893

Country of ref document: US

REEP Request for entry into the european phase

Ref document number: 2014841969

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2014841969

Country of ref document: EP