TW201608525A - System and method for obtaining a motion zone via image vector analysis - Google Patents

Abstract

A system and a method for obtaining a motion zone via image vector analysis comprise an image capturing unit and a processing unit. The image capturing unit captures several continuous images. The processing unit comprises a motion vector analysis module, a background determination module, and a control signal generation module. The motion vector analysis module obtains a vector variation value of the continuous images, the background determination module obtains a background image according to the vector variation value, and the control signal generation module generates a control signal according to the vector variation value to control the image capturing unit.

Description

System and method for obtaining a motion region through image vector analysis

The present invention relates to a system and method for obtaining a motion region through image vector analysis, and more particularly to a system and method for analyzing a foreground and background image through an image vector to obtain a motion region.

The image monitoring system can be used in the management of home, community environment, workplace and public places to monitor the entry and exit of personnel to achieve remote control functions such as security or access control. Security monitoring is an application type of image monitoring system. Through the image monitoring system, managers can instantly monitor the inbound and outbound personnel in a specific area to prevent outsiders from invading.

The general image monitoring system is composed of front-end photography equipment (such as camera), transmission equipment (such as cable, wireless transmission module or network communication module, etc.) and back-end control display equipment (such as computer mainframe and display). composition. The front-end photography device captures the image information, transmits the image information to the back-end control display device through the transmission device for screen display, and the management personnel monitors the display screen through the back-end control display device, controls the front-end monitoring device, and performs image storage on the image information. Wait for processing.

The above image monitoring system needs to control the display device through the back end. The front-end monitoring device can be controlled and image processing is completed. Not only does the system hardware occupy a lot of space, but the system architecture is complicated and difficult to maintain, and the image capturing angle is not easy to adjust, so that most people cannot experience the benefits brought by this technology.

The present invention provides a system and method for obtaining a motion region through image vector analysis, and uses image processing technology to detect a place that changes frequently on a screen to improve the convenience of a block that is easily moved by a person such as a channel. .

The invention provides a system for obtaining a motion region through image vector analysis, comprising an image capturing unit and a processing unit. The image capture unit obtains a plurality of consecutive images, and the processing unit includes a motion vector analysis module, a background determination module, and a control signal generation module, wherein the motion vector analysis module obtains a vector change of the continuous images. The background determining module obtains a background image according to the vector change value, and the control signal generating module generates a control signal according to the vector change value to control the image capturing unit.

In an embodiment of the present invention, the continuous images are divided into a plurality of image blocks, each of the plurality of image blocks includes an image feature vector and a label, and the image block includes image subtraction techniques to obtain image features. The vector, wherein the background subtraction technique comprises identifying a foreground image information and a background image information in the image block, and further, the image feature vector is operated by an image feature vector, and the image feature vector operation includes foreground image information.

In an embodiment of the invention, the processing unit further includes an image conversion module for converting the continuous images into a binarized image.

The present invention also provides a method for obtaining a motion region by image vector analysis, comprising: obtaining a plurality of consecutive images through an image capturing unit; and acquiring a plurality of consecutive images by using a motion vector analysis module of a processing unit a vector change value; a background judgment module obtains a background image according to the vector change value; compares a current image with the background image to determine whether the vector change value exceeds a preset value; and utilizes a control signal generation module A control signal is generated according to the vector change value to control the image capture unit.

In an embodiment of the present invention, the step of using a motion vector analysis module of a processing unit to obtain a vector change value of the consecutive images further includes: applying a background subtraction technique to obtain the consecutive images. An image feature vector.

In an embodiment of the present invention, the method further includes: identifying, by using a background subtraction technique, a foreground image information and a background image information in the image block, wherein the image feature vector is calculated by using an image feature vector, and the image feature vector operation includes the image feature vector operation Prospect image information.

In an embodiment of the present invention, the step of comparing a current image with the background image to determine whether the vector change value exceeds a preset value further includes: when the vector change value exceeds a preset value A binarized image is generated by an image conversion module.

The invention has the characteristics of automatic learning, and can automatically determine the recognition priority of each area in the image according to the time without any operation and setting by the user, so that the performance of the machine can be concentrated in the area where the event is more easily triggered. Take Focus on the frequency of image recognition in specific areas, improve the monitoring application value of IP Camera, and remind users of potentially dangerous image areas.

110‧‧‧Image capture unit

120‧‧‧Processing unit

130‧‧‧Motion Vector Analysis Module

140‧‧‧Background judgment module

150‧‧‧Image Conversion Module

160‧‧‧Control signal generation module

S210~S250‧‧‧Step process

S310~S320‧‧‧Step process

S410~S430‧‧‧Step process

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a block diagram of the system components of the present invention.

Figure 2 is a flow chart of the steps in accordance with Figure 1.

3 to 4 are detailed flowcharts of Fig. 2.

The above described features and advantages of the present invention will be more apparent from the following description.

Please refer to FIG. 1 to FIG. 4 at the same time. FIG. 1 is a block diagram of the system components of the present invention. Figure 2 is a flow chart of the steps in accordance with Figure 1. 3 to 4 are detailed flowcharts of Fig. 2.

In FIG. 1 , a system for obtaining a motion region by image vector analysis includes an image capture unit 110 and a processing unit 120 . The processing unit 120 includes a motion vector analysis module 130 and a background determination module 140 . An image conversion module 150 and a control signal generation module 160.

The image capturing unit 110 obtains a plurality of consecutive images (step S210), and the motion vector analyzing module 130 obtains a vector change value of the consecutive images (step S220), and the background determining module 140 obtains a background image according to the vector change value. (Step S230).

In step S230, further, a background subtraction technique can be used to identify a foreground image information and a background image information in the image block, wherein the image feature vector is calculated by using an image feature vector, and the image feature vector operation includes the foreground Image information.

Then, comparing the current image with the background image, determining whether the vector change value exceeds a preset value (step S240), and if so, generating a binarized image through the image conversion module 150 (step S241), and according to the The image adjusts the weight of each point. Otherwise, the control signal generating module generates a control signal according to the vector change value to control the image capturing unit (step S250).

In this embodiment, the process of comparing the transmitted images (step S310), determining whether a preset value is exceeded (step S320), and generating a binarized image (step S330) may be used in each successive image. The plurality of image blocks include an image feature vector and a landmark as a basis for adjustment.

Furthermore, the image block includes applying a background subtraction technique to obtain an image feature vector, wherein the background subtraction technique includes identifying a foreground image information and a background image information in the image block, and the foreground image information refers to a vector change value. For the block portion that exceeds a preset value, the background image information refers to the block portion where the vector change value does not exceed a preset value. Therefore, when the vector variation value of the background image information exceeds the preset value, the generated binarized image must be used as the basis for the adjustment.

As described above, in FIG. 4, when it is determined whether or not it is a stationary background (step S410), that is, based on the vector change value of the background image information, when the vector change value is smaller than the preset value, it is judged as YES, and Detecting moving objects (steps) Step S420), when the moving object passes, the vector change value is greater than the preset value, that is, the foreground image information, and finally the channel portion through which the moving object passes is listed as the warning area (step S430), so that the control signal generating module The generated control signal controls the image capturing unit to enhance the recording of the warning area.

In summary, the present invention has the feature of automatic learning, and can automatically determine the recognition priority of each area in the image according to time without any operation and setting by the user, so that the performance of the machine can be concentrated on the easier trigger event. The area is used to focus the recognition frequency of specific areas, improve the monitoring application value of IP Camera, and remind users of potentially dangerous image areas.

While the present invention has been described above in the foregoing embodiments, it is not intended to limit the invention, and the equivalents of the modifications and retouchings are still in the present invention without departing from the spirit and scope of the invention. Within the scope of patent protection.

110‧‧‧Image capture unit

120‧‧‧Processing unit

130‧‧‧Motion Vector Analysis Module

140‧‧‧Background judgment module

150‧‧‧Image Conversion Module

160‧‧‧Control signal generation module

Claims (10)

A system for obtaining a motion region by image vector analysis, comprising: an image capturing unit for acquiring a plurality of consecutive images; and a processing unit comprising: a motion vector analysis module for acquiring a vector of the continuous images a change value; a background determining module, obtaining a background image according to the vector change value; and a control signal generating module, generating a control signal according to the vector change value to control the image capturing unit. A system for obtaining a motion region by image vector analysis as described in claim 1, wherein the continuous images are divided into a plurality of image blocks, each of the plurality of image blocks comprising an image feature vector and a landmark. A system for obtaining a motion region by image vector analysis as described in claim 2, wherein the image block includes applying a background subtraction technique to obtain the image feature vector. A system for obtaining a motion region by image vector analysis as described in claim 3, wherein the background subtraction technique comprises identifying a foreground image information and a background image information in the image block. A system for obtaining a motion region by image vector analysis as described in claim 2, wherein the image feature vector is computed using an image feature vector operation, the image feature vector operation including the foreground image information. A system for obtaining a motion region by image vector analysis as described in claim 1, wherein the processing unit further includes an image conversion module to These continuous images are converted into a binarized image. A method for obtaining a motion region by image vector analysis, comprising: obtaining a plurality of consecutive images through an image capturing unit; and using a motion vector analysis module of a processing unit to obtain a vector variation value of the consecutive images; The vector variation value causes a background determination module to obtain a background image; comparing a current image with the background image to determine whether the vector variation value exceeds a preset value; and using a control signal generation module to generate the image according to the vector The change value produces a control signal to control the image capture unit. The method for obtaining a motion region by image vector analysis according to claim 7 of the patent application, wherein the step of acquiring a vector variation value of the consecutive images by using a motion vector analysis module of a processing unit is further The method includes: applying a background subtraction technique to obtain an image feature vector of the continuous images. The method for obtaining a motion region by image vector analysis as described in claim 8 further includes: using the background subtraction technique to identify a foreground image information and a background image information in the image block, wherein the image The feature vector is computed using an image feature vector operation that includes the foreground image information. A method for obtaining a motion region by image vector analysis as described in claim 6 wherein the current image is compared with the background image to determine whether the vector variation value exceeds a preset value. , including: When the vector change value exceeds a preset value, a binarized image is generated through an image conversion module.