WO2012081969A1 - A system and method to detect intrusion event - Google Patents

A system and method to detect intrusion event Download PDF

Info

Publication number
WO2012081969A1
WO2012081969A1 PCT/MY2011/000164 MY2011000164W WO2012081969A1 WO 2012081969 A1 WO2012081969 A1 WO 2012081969A1 MY 2011000164 W MY2011000164 W MY 2011000164W WO 2012081969 A1 WO2012081969 A1 WO 2012081969A1
Authority
WO
WIPO (PCT)
Prior art keywords
blob
properties
motion
pixels
component
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/MY2011/000164
Other languages
French (fr)
Inventor
Kadim Zulaikha
Sze Ling Tang
Kim Meng Liang
Samudin Norshuhada
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mimos Bhd
Original Assignee
Mimos Bhd
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 Mimos Bhd filed Critical Mimos Bhd
Publication of WO2012081969A1 publication Critical patent/WO2012081969A1/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19604—Image analysis to detect motion of the intruder, e.g. by frame subtraction involving reference image or background adaptation with time to compensate for changing conditions, e.g. reference image update on detection of light level change
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/174—Segmentation; Edge detection involving the use of two or more images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/254—Analysis of motion involving subtraction of images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19606—Discriminating between target movement or movement in an area of interest and other non-signicative movements, e.g. target movements induced by camera shake or movements of pets, falling leaves, rotating fan
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30232—Surveillance

Definitions

  • a SYSTEM AND METHOD TO DETECT INTRUSION EVENT FIELD OF INVENTION The present invention relates to a system and method for detection of intrusion event within one or more predefined regions of interest (ROIs) while minimizing false alarm due to noise and reduction of misdetection.
  • ROIs regions of interest
  • video surveillance utilizes human operators to monitor multiple screens and to identify occurrence of any abnormal event in monitored area.
  • Existing intrusion detection methods are based on motion detection technology, whereby existence of object within the scene is detected by frame differencing current image frame with reference image which is generated from series of history images.
  • this area of intrusion detection refers to 'blob detection' where blobs are visual modules.
  • the detection method is aimed at detecting points and/or regions in the image that are either brighter or darker than the surrounding area or ROI.
  • the present invention provides a system and method to detect intrusion event within one or more predefined regions-of-interest by validating noise removal analysis prior to actual removal, updating background image by considering object significance level, and including temporal analysis during event analysis based on a predefined sensitivity level for each region-of-interest in the scene.
  • the subject matter claimed herein is not limited to embodiments that solve any disadvantages or that operate only in environments such as those described above. Rather, this background is only provided to illustrate one exemplary technology area where some embodiments described herein may be practice.
  • the present invention provides a system (100) for detecting intrusion event in region of interest.
  • the system comprising at least one blob segmentation sub-component (102), at least one blob properties computation sub-component (104), at least one blob filtering analysis and validation sub-component (106), at least one background model update sub-component (108) and at least one intrusion event analysis sub-component (1 0).
  • the said blob segmentation sub-component (102) processes current image and extracts motion pixels by subtracting current image from reference or background image.
  • the said blob properties computation sub-component (104) classifies motion pixels by applying morphology closing and filling holes to current motion map and labeling each of adjacent motion pixels.
  • the said blob filtering analysis and validation sub-component (106) validates intruder blobs using combination of rules based on blob geometric properties, temporal properties and blob significancy based on similarity properties of blob with surrounding pixels.
  • the said background model update sub-component (108) updates background image using combination of rules based on similarity value of detected blob and its surrounding by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant while the said intrusion event analysis sub-component (110) triggers alert when event is detected and verified.
  • the said blob properties computation sub-component (104) determines intruder blobs by determining similarity between blob and surrounding pixels in current image by calculating intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image, determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image and determining motion blob as non-siginificant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value.
  • the said blob filtering analysis and validation sub-component (106) calculates blob properties validates intruder blobs using rule set which combines information of calculated properties by considering blob as noise if blob status is new and blob is non- significant, examining blob geometric properties against acceptable range of values if blob status is not new, considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value else considering blob as intruder blob.
  • Another aspect of the present invention provides a method (700) for detecting intrusion event in region of interest.
  • the method comprising the steps of processing current image and extracting motion pixels (702), classifying motion pixels (704), determining intruder blobs using combination of rules based on blob geometric properties, temporal properties and similarity properties of blob with surrounding pixels (706), updating background image using combination of rules based on similarity value of detected blob and its surrounding (708) and triggering alert when event is detected and verified (710).
  • a method for determining intruder blobs further comprises calculating blob properties (802) and validating intruder blobs using rule set which combines information of calculated properties (804).
  • a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding further comprises updating pixel in background image using predefined increment or decrement values if current pixel is background pixel (902) and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant (904).
  • a method for calculating blob properties further comprises calculating blob geometrical properties (1002), calculating blob temporal properties (1004) and calculating blob significancy based on similarity properties with surrounding pixels (1006).
  • a method for determining intruder blobs further comprises determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image (1102), determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image (1104) and determining motion blob as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value ( 106).
  • a method for validating intruder blobs using rule set which combines information of calculated properties further comprises considering blob as noise if blob status is new and blob is non-significant (1202), examining blob geometric properties against acceptable range of values if blob status is not new (1204), considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value (1206), else considering blob as intruder blob (1208).
  • Fig. 1 illustrates architecture of intrusion detection system of the present invention.
  • Fig. 2 illustrates blob segmentation sub-component.
  • Fig. 3 is a flowchart illustrating blob filtering analysis and validation sub-component.
  • Fig. 4 is a flowchart illustrating blob properties analysis.
  • Fig. 5 is a flowchart illustrating motion and non-motion histogram intersection computation.
  • Fig. 6 illustrates background updating rule.
  • Fig. 7 is a flowchart illustrating a method for detecting intrusion event in region of interest.
  • Fig. 8 is a flowchart illustrating a method for determining intruder blobs.
  • Fig. 9 is a flowchart illustrating a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding.
  • Fig. 10 is a flowchart illustrating a method for calculating blob properties.
  • Fig. 11 is a flowchart illustrating a method for further determining intruder blob.
  • Fig. 12 is a flowchart illustrating a method for validating intruder blobs using rule set which combines information of calculated properties.
  • the present invention provides a system and method for detection of intrusion event within one or more predefined regions of interest (ROIs) while minimizing false alarm due to noise and reduction of misdetection.
  • ROIs regions of interest
  • Fig. 1 illustrates general architecture of the present invention for intrusion event detection.
  • the illustrated intrusion detection system (100) comprises at least one blob segmentation sub-component (102), at least one blob properties computation sub-component (104), at least one blob filtering analysis and validation sub-component (106), at least one background model update sub-component (108) and at least one intrusion event analysis sub-component (110).
  • the said system comprises of inputs wherein the inputs includes at least one sensor device, at least one region of interest (ROI) map and at least one sensitivity level detection.
  • Sensitivity level determines responsiveness of the system of the present invention by detecting intrusion event as well as to provide for consistency of the system in maintaining event alert.
  • blob segmentation subcomponent (102) processes current image and extracts motion pixels by subtracting current image from reference or background image. The difference of the image is thresholded to produce current motion map.
  • Motion map is a binary map where pixels with zero values indicate background pixels whereas non-zero pixels are motion or foreground pixels.
  • Blob properties computation sub-component (104) classifies motion pixels by applying morphology closing and filling holes to current motion map. Resultant map is labeled using pixel connect algorithm whereby each group of connected motion pixels will be labeled with the same label to indicate pixels which belongs to the same group. A connected motion pixel is called motion blob.
  • the said blob properties computation sub-component (104) further determines blobs significancy by determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image, determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image and determining motion blob as nonsignificant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value.
  • Fig. 3 is a flowchart illustrating blob filtering analysis and validation sub-component (106).
  • Blob filtering analysis and validation subcomponent (106) validates intruder blobs using rule set which combines information of calculated properties based on blob geometrical properties, temporal properties and blob significancy based on similarity properties of blob with surrounding pixels.
  • Blob filtering analysis and validation sub-component (106) validates intruder blobs using rule set which combines information of calculated properties by considering blob as noise if blob status is new and blob is non-significant, examining blob geometric properties against acceptable range of values if blob status is not new, considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value else considering blob as intruder blob.
  • process flow of blob filtering analysis and validation subcomponent (106) comprises of current blobs properties analysis, blobs temporal analysis, blobs neighboring pixels analysis and validation of blobs.
  • current blobs properties analysis probability that motion blob belongs to noise based on its properties is computed.
  • User can select to invoke any of the filters, for example size filter, compact filter, or orientation filter, then the properties for each blob will be compared against each filters predetermined minimum and maximum acceptable range of values. Blob will be flagged as 'to be-removed' if any blob properties does not fall within the acceptable range. Prior to this step, all small motion blobs will be removed first. This entire step is illustrated in Fig. 4.
  • blob temporal properties are updated based on current- previous blob overlapping analysis.
  • Each detected blob in current image frame will be compared against all of previous motion blobs and percentage of overlapping between current and previous blob will be calculated. If overlapping percentage is above a certain threshold value, the corresponding blobs are considered as related and vice versa.
  • temporal properties of each detected blob in current image frame will be updated. These properties include blob lifespan, blob's average and standard deviation area, centroid etc.
  • blobs neighboring pixels analysis all motion and non-motion pixels are enclosed within blob bounding box which will be analyzed using histogram intersection in order to determine if motion blob is an actual object or a 'ghost' of object. Thereafter, blobs are validated to be removed and object map is updated.
  • Fig. 5 is a flowchart illustrating motion and non-motion histogram intersection computation of background model update sub-component (108).
  • Color histogram of background pixels and foreground pixels will be computed as illustrated in FIG. 5.
  • background or reference image will be updated. Updating image is crucial as gradual background changes in the scene will be gradually updated as part of background.
  • background updating step will be taken into consideration as the current object significance level which have been computed in previous sub-component.
  • the said background model update sub-component updates background image using combination of rules based on similarity value of detected blob and its surrounding by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant.
  • Fig. 6 Rules to update the said background is illustrated in Fig. 6. For each image pixels, if pixel is background, then update pixel. If pixel is foreground and non-significant then update pixel. Otherwise, pixels will not be updated. After all steps above completed, the final object map will contain only possible human blobs. Then these blobs will be analyzed by event analysis sub-component. The said intrusion event analysis sub-component triggers alert when event is detected and verified. First each of these blobs reference points will be examined whether they fall within the predefined ROI or not. Only blobs which reference points fall within ROI will be considered.
  • Fig. 7 is a flowchart illustrating a method for detecting intrusion event in region of interest.
  • a method (700) for detecting intrusion event in region of interest comprising the steps of processing current image and extracting motion pixels (702), classifying motion pixels (704) by applying morphology closing and filing holes to current motion map and labeling each of adjacent motion pixels, determining intruder blobs using combination of rules based on blob geometric properties, temporal properties and similarity properties of blob with surrounding pixels (706), updating background image using combination of rules based on similarity value of detected blob and its surrounding (708) and triggering alert when event is detected and verified (710).
  • Fig. 8 is a flowchart illustrating a method for determining intruder blobs
  • Fig. 9 is a flowchart illustrating a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding
  • Fig. 10 is a flowchart illustrating a method for calculating blob properties.
  • blob properties is first calculated to determine intruder blobs (802) and thereafter intruder blobs is validated using rule set which combines information of calculated properties (804).
  • rule set which combines information of calculated properties (804).
  • the method for computing blob properties further comprises calculating blob geometrical properties (1002), calculating blob temporal properties (1004) and calculating blob significancy based on similarity properties with surrounding pixels (1006).
  • Blob geometrical properties include area, orientation and compactness
  • blob temporal properties include blob lifespan and blob status.
  • background image is updated by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel (902) and background image is updated with current pixel intensity if current pixel belongs to motion blob and blob is non-significant (904).
  • Fig. 11 provides further steps for determining intruder blobs by determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image (1102). Thereafter, similarity between blob and surrounding pixels in background image is determined by calculating histogram intersection value between histograms of motion pixels and non- motion pixels within blob bounding box for background image (1104). Subsequently, motion blob is determined as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value (1106).
  • Fig. 12 a flowchart illustrating a method for validating intruder blobs using rule set which combines information of calculated properties.
  • intruder blobs are validated using rule set which combines information of calculated properties.
  • Blob is considered as noise if blob status is new (1202) and blob is non-significant while blob geometric properties are examined against acceptable range of values if blob status is not new (1204).
  • blob is considered as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value (1206); else blob is considered as intruder blob (1208).
  • Alert is triggered when event is detected and verified wherein only blobs which reference points fall within ROI will be considered.
  • the said intrusion event analysis subcomponent (110) triggers alert when event is detected and verified.
  • the approach of the present invention provides a system and method to detect intrusion event within one or more predefined regions-of-interest by validating noise removal analysis prior to actual removal, updating background image by considering object significance level, and including temporal analysis during event analysis based on a predefined sensitivity level for each region-of-interest in the scene.
  • the present invention may be embodied in other specific forms without departing from its essential characteristics.
  • the described embodiments are to be considered in all respects only as illustrative and not restrictive.
  • the scope of the invention is, therefore indicated by the appended claims rather than by the foregoing description. All changes, which come within the meaning and range of equivalency of the claims, are to be embraced within their scope.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

The present invention provides a system and method for detecting intrusion event in region of interest. The system comprises at least one blob segmentation sub-component which processes current image and extracts motion pixels by subtracting current image from reference or background image; at least one blob properties computation sub¬ component which classifies motion pixels by applying morphology closing and filing holes to current motion map and labelling each of adjacent motion pixels; at least one blob filtering analysis and validation sub-component which validates intruder blobs using rule sets; at least one background model update sub-component which updates background image using combination of rules based on similarity value of detected blob and its surrounding; and at least one intrusion event analysis sub-component which triggers alert when event is detected and verified.

Description

A SYSTEM AND METHOD TO DETECT INTRUSION EVENT FIELD OF INVENTION The present invention relates to a system and method for detection of intrusion event within one or more predefined regions of interest (ROIs) while minimizing false alarm due to noise and reduction of misdetection.
BACKGROUND ART
Conventionally, video surveillance utilizes human operators to monitor multiple screens and to identify occurrence of any abnormal event in monitored area. Existing intrusion detection methods are based on motion detection technology, whereby existence of object within the scene is detected by frame differencing current image frame with reference image which is generated from series of history images.
Accuracy of existing intrusion detection methods has a high occurence of false detection is present due to noise. There is also a low detection rate due to filtering of object of interest by noise removal and false detection due to ghost effect. Specifically, this area of intrusion detection refers to 'blob detection' where blobs are visual modules. The detection method is aimed at detecting points and/or regions in the image that are either brighter or darker than the surrounding area or ROI.
The present invention provides a system and method to detect intrusion event within one or more predefined regions-of-interest by validating noise removal analysis prior to actual removal, updating background image by considering object significance level, and including temporal analysis during event analysis based on a predefined sensitivity level for each region-of-interest in the scene. The subject matter claimed herein is not limited to embodiments that solve any disadvantages or that operate only in environments such as those described above. Rather, this background is only provided to illustrate one exemplary technology area where some embodiments described herein may be practice. SUMMARY OF INVENTION
The present invention provides a system (100) for detecting intrusion event in region of interest. The system comprising at least one blob segmentation sub-component (102), at least one blob properties computation sub-component (104), at least one blob filtering analysis and validation sub-component (106), at least one background model update sub-component (108) and at least one intrusion event analysis sub-component (1 0). The said blob segmentation sub-component (102) processes current image and extracts motion pixels by subtracting current image from reference or background image. The said blob properties computation sub-component (104) classifies motion pixels by applying morphology closing and filling holes to current motion map and labeling each of adjacent motion pixels. The said blob filtering analysis and validation sub-component (106) validates intruder blobs using combination of rules based on blob geometric properties, temporal properties and blob significancy based on similarity properties of blob with surrounding pixels. The said background model update sub-component (108) updates background image using combination of rules based on similarity value of detected blob and its surrounding by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant while the said intrusion event analysis sub-component (110) triggers alert when event is detected and verified.
Further, the said blob properties computation sub-component (104) determines intruder blobs by determining similarity between blob and surrounding pixels in current image by calculating intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image, determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image and determining motion blob as non-siginificant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value.
Preferably, the said blob filtering analysis and validation sub-component (106) calculates blob properties validates intruder blobs using rule set which combines information of calculated properties by considering blob as noise if blob status is new and blob is non- significant, examining blob geometric properties against acceptable range of values if blob status is not new, considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value else considering blob as intruder blob.
Another aspect of the present invention provides a method (700) for detecting intrusion event in region of interest. The method comprising the steps of processing current image and extracting motion pixels (702), classifying motion pixels (704), determining intruder blobs using combination of rules based on blob geometric properties, temporal properties and similarity properties of blob with surrounding pixels (706), updating background image using combination of rules based on similarity value of detected blob and its surrounding (708) and triggering alert when event is detected and verified (710).
Preferably, a method for determining intruder blobs, further comprises calculating blob properties (802) and validating intruder blobs using rule set which combines information of calculated properties (804).
Further, a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding, further comprises updating pixel in background image using predefined increment or decrement values if current pixel is background pixel (902) and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant (904). A method for calculating blob properties further comprises calculating blob geometrical properties (1002), calculating blob temporal properties (1004) and calculating blob significancy based on similarity properties with surrounding pixels (1006). A method for determining intruder blobs further comprises determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image (1102), determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image (1104) and determining motion blob as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value ( 106). Preferably, a method for validating intruder blobs using rule set which combines information of calculated properties further comprises considering blob as noise if blob status is new and blob is non-significant (1202), examining blob geometric properties against acceptable range of values if blob status is not new (1204), considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value (1206), else considering blob as intruder blob (1208).
The present invention consists of features and a combination of parts hereinafter fully described and illustrated in the accompanying drawings, it being understood that various changes in the details may be made without departing from the scope of the invention or sacrificing any of the advantages of the present invention.
BRIEF DESCRIPTION OF ACCOMPANYING DRAWINGS
To further clarify various aspects of some embodiments of the present invention, a more particular description of the invention will be rendered by references to specific embodiments thereof, which are illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail through the accompanying drawings in which:
Fig. 1 illustrates architecture of intrusion detection system of the present invention.
Fig. 2 illustrates blob segmentation sub-component.
Fig. 3 is a flowchart illustrating blob filtering analysis and validation sub-component. Fig. 4 is a flowchart illustrating blob properties analysis.
Fig. 5 is a flowchart illustrating motion and non-motion histogram intersection computation. Fig. 6 illustrates background updating rule.
Fig. 7 is a flowchart illustrating a method for detecting intrusion event in region of interest. Fig. 8 is a flowchart illustrating a method for determining intruder blobs.
Fig. 9 is a flowchart illustrating a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding. Fig. 10 is a flowchart illustrating a method for calculating blob properties.
Fig. 11 is a flowchart illustrating a method for further determining intruder blob. Fig. 12 is a flowchart illustrating a method for validating intruder blobs using rule set which combines information of calculated properties.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention provides a system and method for detection of intrusion event within one or more predefined regions of interest (ROIs) while minimizing false alarm due to noise and reduction of misdetection. Hereinafter, this specification will describe the present invention according to the preferred embodiments. It is to be understood that limiting the description to the preferred embodiments of the invention is merely to facilitate discussion of the present invention and it is envisioned without departing from the scope of the appended claims.
Reference is first being made to Fig 1. Fig. 1 illustrates general architecture of the present invention for intrusion event detection. The illustrated intrusion detection system (100) comprises at least one blob segmentation sub-component (102), at least one blob properties computation sub-component (104), at least one blob filtering analysis and validation sub-component (106), at least one background model update sub-component (108) and at least one intrusion event analysis sub-component (110). The said system comprises of inputs wherein the inputs includes at least one sensor device, at least one region of interest (ROI) map and at least one sensitivity level detection. Sensitivity level determines responsiveness of the system of the present invention by detecting intrusion event as well as to provide for consistency of the system in maintaining event alert.
Reference is now being made to Fig 2. As illustrated in Fig. 2, blob segmentation subcomponent (102) processes current image and extracts motion pixels by subtracting current image from reference or background image. The difference of the image is thresholded to produce current motion map. Motion map is a binary map where pixels with zero values indicate background pixels whereas non-zero pixels are motion or foreground pixels.
Blob properties computation sub-component (104) classifies motion pixels by applying morphology closing and filling holes to current motion map. Resultant map is labeled using pixel connect algorithm whereby each group of connected motion pixels will be labeled with the same label to indicate pixels which belongs to the same group. A connected motion pixel is called motion blob. The said blob properties computation sub-component (104) further determines blobs significancy by determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image, determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image and determining motion blob as nonsignificant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value.
For each motion blob, its geometric properties such as blob area, orientation, compactness, projection, skewness etc. will be determined.
Reference is now being made to Fig. 3. Fig. 3 is a flowchart illustrating blob filtering analysis and validation sub-component (106). Blob filtering analysis and validation subcomponent (106) validates intruder blobs using rule set which combines information of calculated properties based on blob geometrical properties, temporal properties and blob significancy based on similarity properties of blob with surrounding pixels. Blob filtering analysis and validation sub-component (106) validates intruder blobs using rule set which combines information of calculated properties by considering blob as noise if blob status is new and blob is non-significant, examining blob geometric properties against acceptable range of values if blob status is not new, considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value else considering blob as intruder blob.
As illustrated in Fig. 3, process flow of blob filtering analysis and validation subcomponent (106) comprises of current blobs properties analysis, blobs temporal analysis, blobs neighboring pixels analysis and validation of blobs. In current blobs properties analysis, probability that motion blob belongs to noise based on its properties is computed. User can select to invoke any of the filters, for example size filter, compact filter, or orientation filter, then the properties for each blob will be compared against each filters predetermined minimum and maximum acceptable range of values. Blob will be flagged as 'to be-removed' if any blob properties does not fall within the acceptable range. Prior to this step, all small motion blobs will be removed first. This entire step is illustrated in Fig. 4.
In blobs temporal analysis, blob temporal properties are updated based on current- previous blob overlapping analysis. Each detected blob in current image frame will be compared against all of previous motion blobs and percentage of overlapping between current and previous blob will be calculated. If overlapping percentage is above a certain threshold value, the corresponding blobs are considered as related and vice versa. From the overlapping analysis, temporal properties of each detected blob in current image frame will be updated. These properties include blob lifespan, blob's average and standard deviation area, centroid etc.
In blobs neighboring pixels analysis, all motion and non-motion pixels are enclosed within blob bounding box which will be analyzed using histogram intersection in order to determine if motion blob is an actual object or a 'ghost' of object. Thereafter, blobs are validated to be removed and object map is updated.
Reference is now made to Fig 5. Fig. 5 is a flowchart illustrating motion and non-motion histogram intersection computation of background model update sub-component (108). Color histogram of background pixels and foreground pixels will be computed as illustrated in FIG. 5. In this sub-component, background or reference image will be updated. Updating image is crucial as gradual background changes in the scene will be gradually updated as part of background. To eliminate updating static object-of-interest as background and mistakenly detect 'ghost' of the object-of-interest as a valid object, background updating step will be taken into consideration as the current object significance level which have been computed in previous sub-component. The said background model update sub-component updates background image using combination of rules based on similarity value of detected blob and its surrounding by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant.
Rules to update the said background is illustrated in Fig. 6. For each image pixels, if pixel is background, then update pixel. If pixel is foreground and non-significant then update pixel. Otherwise, pixels will not be updated. After all steps above completed, the final object map will contain only possible human blobs. Then these blobs will be analyzed by event analysis sub-component. The said intrusion event analysis sub-component triggers alert when event is detected and verified. First each of these blobs reference points will be examined whether they fall within the predefined ROI or not. Only blobs which reference points fall within ROI will be considered.
Reference is now being made to Fig. 7. Fig. 7 is a flowchart illustrating a method for detecting intrusion event in region of interest. As illustrated in Fig. 7, a method (700) for detecting intrusion event in region of interest, comprising the steps of processing current image and extracting motion pixels (702), classifying motion pixels (704) by applying morphology closing and filing holes to current motion map and labeling each of adjacent motion pixels, determining intruder blobs using combination of rules based on blob geometric properties, temporal properties and similarity properties of blob with surrounding pixels (706), updating background image using combination of rules based on similarity value of detected blob and its surrounding (708) and triggering alert when event is detected and verified (710).
Reference is now being made to Figs. 8, 9, and 10 respectively. Fig. 8 is a flowchart illustrating a method for determining intruder blobs while Fig. 9 is a flowchart illustrating a method for updating background image using combination of rules based on similarity value of detected blob and its surrounding and Fig. 10 is a flowchart illustrating a method for calculating blob properties. As illustrated in Fig. 8, blob properties is first calculated to determine intruder blobs (802) and thereafter intruder blobs is validated using rule set which combines information of calculated properties (804). As illustrated in Fig. 10, the method for computing blob properties further comprises calculating blob geometrical properties (1002), calculating blob temporal properties (1004) and calculating blob significancy based on similarity properties with surrounding pixels (1006). Blob geometrical properties include area, orientation and compactness while blob temporal properties include blob lifespan and blob status. As illustrated in Fig. 9, background image is updated by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel (902) and background image is updated with current pixel intensity if current pixel belongs to motion blob and blob is non-significant (904).
Reference is now being made to Fig. 11. Fig. 11 provides further steps for determining intruder blobs by determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image (1102). Thereafter, similarity between blob and surrounding pixels in background image is determined by calculating histogram intersection value between histograms of motion pixels and non- motion pixels within blob bounding box for background image (1104). Subsequently, motion blob is determined as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value (1106).
Reference is now being made to Fig. 12. Fig. 12 a flowchart illustrating a method for validating intruder blobs using rule set which combines information of calculated properties. As illustrated in Fig. 12, intruder blobs are validated using rule set which combines information of calculated properties. Blob is considered as noise if blob status is new (1202) and blob is non-significant while blob geometric properties are examined against acceptable range of values if blob status is not new (1204). Thereafter, blob is considered as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value (1206); else blob is considered as intruder blob (1208).
Alert is triggered when event is detected and verified wherein only blobs which reference points fall within ROI will be considered. The said intrusion event analysis subcomponent (110) triggers alert when event is detected and verified.
The approach of the present invention provides a system and method to detect intrusion event within one or more predefined regions-of-interest by validating noise removal analysis prior to actual removal, updating background image by considering object significance level, and including temporal analysis during event analysis based on a predefined sensitivity level for each region-of-interest in the scene. The present invention may be embodied in other specific forms without departing from its essential characteristics. The described embodiments are to be considered in all respects only as illustrative and not restrictive. The scope of the invention is, therefore indicated by the appended claims rather than by the foregoing description. All changes, which come within the meaning and range of equivalency of the claims, are to be embraced within their scope.

Claims

O 2012/081969 13 CLAIMS
1. A system (100) for detecting intrusion event in region of interest, comprising: at least one blob segmentation sub-component (102);
at least one blob properties computation sub-component (104);
at least one blob filtering analysis and validation sub-component (106); at least one background model update sub-component (108); and at least one intrusion event analysis sub-component (110).
2. A system (100) according to Claim 1 , comprises of inputs to the system, the said inputs includes at least one sensor device; at least one region of interest (ROI) map; and at least one sensitivity level detection.
3. A system (100) according to Claim 1 , wherein the said blob segmentation sub-component (102) processes current image and extracts motion pixels by subtracting current image from reference or background image.
4. A system (100) according to Claim 1 , wherein the said blob properties computation sub-component (104) classifies motion pixels by applying morphology closing and filing holes to current motion map and labeling each of adjacent motion pixels.
5. A system (100) according to Claim 1 , wherein the said blob filtering analysis and validation sub-component (106) validates intruder blobs using combination of rules based on blob geometric properties, temporal properties and blob significancy based on similarity properties of blob with surrounding pixels.
6. A system (100) according to Claim 1 , wherein the said background model update sub-component (108) updates background image using combination of rules based on similarity value of detected blob and its surrounding by updating pixel in background image using predefined increment or decrement values if current pixel is background pixel and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non- significant.
A system (100) according to Claim 1 , wherein the said intrusion event analysis sub-component (110) triggers alert when event is detected and verified.
8. A system (100) according to Claim 1 and Claim 4 wherein the said blob properties computation sub-component (104) calculates blobs significancy by:
determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image; and
determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image;
determining motion blob as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value.
9. A system (100) according to Claim 1 and Claim 4, wherein the said blob filtering analysis and validation sub-component (106) validates intruder blobs using rule set which combines information of calculated properties by:
considering blob as noise if blob status is new and blob is non-significant; examining blob geometric properties against acceptable range of values if blob status is not new;
considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value; else
considering blob as intruder blob.
10. A method (700) for detecting intrusion event in region of interest, comprising the steps of:
processing current image and extracting motion pixels (702);
classifying motion pixels (704);
determining intruder blobs using combination of rules based on blob geometric properties, temporal properties and similarity properties of blob with surrounding pixels (706);
updating background image using combination of rules based on similarity value of detected blob and its surrounding (708); and
triggering alert when event is detected and verified (710).
11. A method according to Claim 10, wherein classifying motion pixels further comprises applying morphology closing and filing holes to current motion map and labeling each of adjacent motion pixels.
12. A method according to Claim 10, wherein determining intruder blobs further comprises:
calculating blob properties (802); and
validating intruder blobs using rule set which combines information of calculated properties (804).
13. A method according to Claim 10, wherein updating background image using combination of rules based on similarity value of detected blob and its surrounding, further comprising:
updating pixel in background image using predefined increment or decrement values if current pixel is background pixel (902); and updating background pixel with current pixel intensity if current pixel belongs to motion blob and blob is non-significant (904).
14. A method according to Claim 12, wherein calculating blob properties further comprises:
calculating blob geometrical properties (1002);
calculating blob temporal properties (1004); and calculating blob significancy based on similarity properties with
surrounding pixels (1006).
15. A method according to Claim 10, wherein determining intruder blobs, further comprises:
determining similarity between blob and surrounding pixels in current image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for current image (1102);
determining similarity between blob and surrounding pixels in background image by calculating histogram intersection value between histograms of motion pixels and non-motion pixels within blob bounding box for background image (1104);
determining motion blob as non-significant if histogram intersection value of background image is larger than histogram intersection value of current image for more than predefined threshold value (1106).
16. A method according to Claim 12, wherein validating intruder blobs using rule set which combines information of calculated properties further comprising: considering blob as noise if blob status is new and blob is non-significant (1202);
examining blob geometric properties against acceptable range of values if blob status is not new (1204);
considering blob as noise if blob property is out of predefined range and blob status is not split and blob lifespan is more than predefined threshold value (1206); else
considering blob as intruder blob (1208).
17. A method according to Claim 14, wherein blob geometrical properties includes area, orientation and compactness.
18. A method according to Claim 14, wherein blob temporal properties includes blob lifespan and blob status.
PCT/MY2011/000164 2010-12-13 2011-06-30 A system and method to detect intrusion event Ceased WO2012081969A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
MYPI2010005928 2010-12-13
MYPI2010005928 MY152782A (en) 2010-12-13 2010-12-13 A system and method to detect intrusion event

Publications (1)

Publication Number Publication Date
WO2012081969A1 true WO2012081969A1 (en) 2012-06-21

Family

ID=46244924

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/MY2011/000164 Ceased WO2012081969A1 (en) 2010-12-13 2011-06-30 A system and method to detect intrusion event

Country Status (2)

Country Link
MY (1) MY152782A (en)
WO (1) WO2012081969A1 (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103209321A (en) * 2013-04-03 2013-07-17 南京邮电大学 Method for quickly updating video background
CN103606166A (en) * 2013-12-04 2014-02-26 天津普达软件技术有限公司 Pill filling image segmentation method
WO2017129804A1 (en) * 2016-01-29 2017-08-03 Kiwisecurity Software Gmbh Methods and apparatus for using video analytics to detect regions for privacy protection within images from moving cameras
CN110244998A (en) * 2019-06-13 2019-09-17 广州酷狗计算机科技有限公司 Page layout background, the setting method of live page background, device and storage medium
CN113158800A (en) * 2021-03-19 2021-07-23 上海云赛智联信息科技有限公司 Enclosure intrusion hybrid detection method and enclosure intrusion hybrid detection system
CN118247933A (en) * 2024-04-18 2024-06-25 江苏臻星美悦健康科技有限公司 On-site dissuasion and warning system using image big data processing

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050100192A1 (en) * 2003-10-09 2005-05-12 Kikuo Fujimura Moving object detection using low illumination depth capable computer vision
US20060140481A1 (en) * 2004-12-08 2006-06-29 Kye-Kyung Kim Target detecting system and method
US20080166045A1 (en) * 2005-03-17 2008-07-10 Li-Qun Xu Method of Tracking Objects in a Video Sequence
US20080219573A1 (en) * 2007-02-28 2008-09-11 Interuniversitair Microelektronica Centrum Vzw (Imec) System and method for motion detection and the use thereof in video coding
WO2010124062A1 (en) * 2009-04-22 2010-10-28 Cernium Corporation System and method for motion detection in a surveillance video

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050100192A1 (en) * 2003-10-09 2005-05-12 Kikuo Fujimura Moving object detection using low illumination depth capable computer vision
US20060140481A1 (en) * 2004-12-08 2006-06-29 Kye-Kyung Kim Target detecting system and method
US20080166045A1 (en) * 2005-03-17 2008-07-10 Li-Qun Xu Method of Tracking Objects in a Video Sequence
US20080219573A1 (en) * 2007-02-28 2008-09-11 Interuniversitair Microelektronica Centrum Vzw (Imec) System and method for motion detection and the use thereof in video coding
WO2010124062A1 (en) * 2009-04-22 2010-10-28 Cernium Corporation System and method for motion detection in a surveillance video

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
TANG S.L. ET AL.: "Hybrid Blob and Particle Filter Tracking Approach for Robust Object Tracking", INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE, ICCS 2010, PROCEDIA COMPUTER SCIENCE, vol. 1, May 2010 (2010-05-01), pages 2549 - 2557, XP027492187, DOI: doi:10.1016/j.procs.2010.04.288 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103209321A (en) * 2013-04-03 2013-07-17 南京邮电大学 Method for quickly updating video background
CN103209321B (en) * 2013-04-03 2016-04-13 南京邮电大学 A kind of video background Rapid Updating
CN103606166A (en) * 2013-12-04 2014-02-26 天津普达软件技术有限公司 Pill filling image segmentation method
WO2017129804A1 (en) * 2016-01-29 2017-08-03 Kiwisecurity Software Gmbh Methods and apparatus for using video analytics to detect regions for privacy protection within images from moving cameras
US10565395B2 (en) 2016-01-29 2020-02-18 Kiwi Security Software GmbH Methods and apparatus for using video analytics to detect regions for privacy protection within images from moving cameras
US12062268B2 (en) 2016-01-29 2024-08-13 Kiwisecurity Software Gmbh Methods and apparatus for using video analytics to detect regions for privacy protection within images from moving cameras
CN110244998A (en) * 2019-06-13 2019-09-17 广州酷狗计算机科技有限公司 Page layout background, the setting method of live page background, device and storage medium
CN113158800A (en) * 2021-03-19 2021-07-23 上海云赛智联信息科技有限公司 Enclosure intrusion hybrid detection method and enclosure intrusion hybrid detection system
CN118247933A (en) * 2024-04-18 2024-06-25 江苏臻星美悦健康科技有限公司 On-site dissuasion and warning system using image big data processing

Also Published As

Publication number Publication date
MY152782A (en) 2014-11-28

Similar Documents

Publication Publication Date Title
CN110516609B (en) Fire disaster video detection and early warning method based on image multi-feature fusion
CN102236947B (en) Flame monitoring method and system based on video camera
CN106056079B (en) A kind of occlusion detection method of image capture device and human face five-sense-organ
JP5518359B2 (en) Smoke detector
Bayona et al. Comparative evaluation of stationary foreground object detection algorithms based on background subtraction techniques
WO2023039781A1 (en) Method for detecting abandoned object, apparatus, electronic device, and storage medium
WO2012081969A1 (en) A system and method to detect intrusion event
CN112818766A (en) High-altitude parabolic detection alarm method and system based on computer vision
CN101968848A (en) Video monitoring method and system and video monitoring alarm system
CN102722700A (en) Method and system for detecting abandoned object in video monitoring
CN101426128B (en) Detection system and method for stolen and lost packet
CN111667419A (en) A moving target ghost elimination method and system based on Vibe algorithm
CN113297926A (en) Behavior detection and identification method and system
CN112927262A (en) Camera lens shielding detection method and system based on video
CN104809742A (en) Article safety detection method in complex scene
TWI493510B (en) Falling down detection method
WO2012074366A2 (en) A system and a method for detecting a loitering event
KR101690050B1 (en) Intelligent video security system
CN109583396A (en) A kind of region prevention method, system and terminal based on CNN two stages human testing
Tang et al. Hybrid blob and particle filter tracking approach for robust object tracking
US10916016B2 (en) Image processing apparatus and method and monitoring system
Christodoulidis et al. Near real-time human silhouette and movement detection in indoor environments using fixed cameras
Murari et al. Algorithms for the automatic identification of MARFEs and UFOs in JET database of visible camera videos
JP4925942B2 (en) Image sensor
CN118381881A (en) Method and system for realizing regional dangerous behavior management and control and computer program medium

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: 11848668

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11848668

Country of ref document: EP

Kind code of ref document: A1