TWI421795B - System and method of image processing based on color information, and method for image categorization using the same - Google Patents

Description

Processing system, processing method and image classification method using image color information

The present invention discloses a comparison processing system and method using image color information, in particular, using chromaticity space conversion, and retaining only color information as a basis for image comparison to eliminate the problem that may cause erroneous recognition.

The Intelligent Video Surveillance (IVS) system is a closed circuit television (CCTV) network, a console monitor and security personnel. In the active development of intelligent video surveillance systems, digital image analysis can be used to image the images, compare and compare image processing processes, find moving images or analyze meaningful information, and then achieve automatic tracking. Judging, assisting security personnel to identify suspicious objects.

In the prior art, there is a monitoring mode in which a closed-circuit television video is used in conjunction with a security personnel. In addition to the security personnel, the security personnel can determine suspicious objects on the spot, usually after viewing a video event, relying on the monitor. Find useful information on the screen.

Then, after digitalization, the traditional closed-circuit television is changed into a digital camera, which makes it easy to realize automatic detection and judgment through digital control, and saves space and cost of image storage.

At present, the technology of security monitoring further realizes intelligent video surveillance. In addition to generating a variety of monitoring applications, it also applies to network transmission technologies, including performing target object detection, tracking, abnormal event judgment, and then triggering alarms. The monitoring work is more powerful and far-reaching.

In the case of security monitoring, one of the most important things is the detection of moving objects. Because the digital video camera is composed of multiple frames, the reference background is established first. Then, the multiple frames of images are continuously compared to obtain the changes generated within a certain time period, so that the events occurring in the monitored area can be analyzed, and then the relevant security personnel are notified through the alarm mechanism.

Therefore, the pixel difference between the images becomes the basis of the image analysis, and the method for identifying the feature of the object for monitoring the camera disclosed in US Patent Publication No. 2008/0031493 (published on Feb. 7, 2008) is incorporated by reference. That is, as shown in the first figure, a method of object feature recognition through a color chart is shown.

In step S101, an object is identified from a series of images in a series of changes, and then a color chart generator is used to generate a color histogram according to the color and brightness information of the object in each frame (step S103). Further, in step S105, the scene and the object are divided into different types according to the color distribution in the color chart (step S105), and then coordinate conversion of the specific color space may be performed in order to highlight the object feature (step S107), and then calculated. The parameters in the color distribution in the various types (step S109) to derive the object features (step S111).

In this example, the change of the continuous scene is obtained through the analysis of the color distribution to obtain the characteristics of the objects in the monitoring scene. After that, the surveillance camera can track according to the characteristics of the object, or other analysis.

Since the technology of general intelligent video surveillance often obtains a reference image by filtering the background, and then uses numerical analysis to find the change situation, generally the change of the color image is used to find out the movement change, which may be caused by the environment and cause a false alarm. Other conventional techniques have been proposed to reduce various possible false positives by simplifying color images, including the use of grayscale images to exclude unwanted colors, and the use of black and white variations to find moving objects. However, grayscale images may still be affected by variations in light brightness and false positives.

In the prior art, the image recorded by the camera may be misjudged due to ambient light and brightness contrast. The processing system, processing method and image classification method using image color information disclosed in the present invention only capture images in the before and after images. The color information in the chromaticity space, especially the chromaticity space in which the original image is located is converted into the color-saturation-intensity chromaticity space (Hue, Saturation, Intensity, HSI space), and only the color information is captured. To exclude other information that may cause misjudgment. This increases the accuracy of the judgment to produce a correct and effective message.

According to a preferred embodiment, the invention discloses a processing method for using image color information, wherein the step is to first capture a continuous image and set a reference image, and actually implement an initial frame that can be started as a reference image, and then perform chroma. Spatial conversion, in particular, transforms the chromaticity space in which the original image is located into a color-saturation-luminance chromaticity space. Then, the color information of an object in the reference image is extracted, the information of saturation and brightness is excluded, and reference information is established.

After that, the step continues to capture subsequent images, and performs chroma space conversion on each frame in the subsequent image, and also converts into a color-saturation-luminance chromaticity space, and then captures color information of the objects in the subsequent images. According to the matching operation between the reference image and the color information of the object in the subsequent image, it is judged whether the criterion for the change of the object is met. After the matching operation is performed, if it is determined to be a match, the subsequent image is continuously captured; but if the judgment is not matched (indicating a change), a message is sent.

In the above steps, the specific event can be judged by whether the comparison is matched. For example, the mismatch may indicate that the object is moving, and the moving path of the object may be drawn according to the mismatch information of the color information of the object in the subsequent image. A mismatch may indicate that the position and angle of the camera was changed, or that the camera was obscured by foreign objects.

Or, you can detect specific objects in a specific space to determine whether they are missing, or whether other objects are added to a specific space. In another embodiment, a trigger message may be sent through a mismatch determination to trigger the camera to start recording an image, or perform other actions, including stopping recording. In another embodiment, by judging that the objects do not match before and after, the event is determined to be in and out, and then the counting is started.

According to another preferred embodiment, the processing system for using image color information disclosed in the present invention includes an image capturing unit that captures a continuous image, and temporarily stores the continuous image through the temporary memory, and then processes the continuous image by the processing unit. Take out a reference image and subsequent images. The storage unit stores the reference image and the subsequent image captured in a period of time, and then processes the image through an image processing module.

The image processing module accesses the image information in the storage unit, and converts the chromaticity space originally in each frame of the continuous image into a color-saturation-luminance chromaticity space (HSI) through the chromaticity space conversion unit. Space), the color feature extraction unit is used to exclude the characteristics of saturation and brightness, and only the color features of the specific object are extracted. The change of the object is judged by the matching operation of the color feature, and the matching operation is used to determine whether the color feature of the object in the reference image and the object in the continuous image match.

Finally, the matching is used to determine the change of the object to determine whether to send a message, such as determining that the object is abnormal, that is, issuing an alert message.

Since the images in the surveillance often cause false alarms due to various uncertain factors, such as using surveillance cameras to monitor the intrusion of specific spaces, when the space changes, the images captured by the camera will change, resulting in Alarm message, but it is difficult to avoid the problem of being misidentified as a suspicious event due to changes in light and shadow. The present invention discloses a comparison processing system and method using image color information, and includes an image classification method, in particular, using a color space transformation (Color Space Transformation) to convert the chromaticity space in which the original image is located, and only retains Color information is used as the basis for image comparison to eliminate problems that may cause misidentification.

Wherein, the chromaticity space in which the image is originally located may be a RGB color space, but the color value changed due to brightness, light and shadow (such as sunlight change) is hidden, and the relevant threshold is The value setting is difficult. However, if only converted to grayscale expression, it is still based on brightness. Therefore, in particular, the present invention converts the chromaticity space in which the original image is located into a Hue (Saturation, Intensity, HSI space), and only captures the Hue information. The variables in which saturation and intensity (brightness) are excluded are used to exclude other information that may cause misjudgment, and the color information is used to increase the accuracy of the image change to produce a correct and effective message.

If the present invention is applied to the field of security monitoring, a schematic diagram of an application embodiment of a processing system using image color information in the second drawing can be referred to. One or more cameras 23 are provided in the building 21. The images captured by the camera 23 can be transmitted to the remote monitoring center 27 through wireless or wired communication. In this example, the wireless base station 25 performs image transmission.

The live image captured by the present invention will be transmitted to the monitoring center 27, which has a computer system for image processing, through object judgment, chromaticity space conversion, and capturing color (Hue) information as a basis for monitoring. .

In addition to performing image processing in the above-mentioned monitoring center 27, another embodiment can further set a computer system with object judgment, chromaticity space conversion and color information capture function on the camera 23, and then transfer the processed information. To the monitoring center 27.

The monitoring applications may include: lens displacement detection, which detects that the camera lens is not moving normally; lens shadow detection is determined by the disappearance of the captured image to determine whether there is foreign matter obscuration; the lost object detection is for a specific object in a specific space. Monitor and determine whether objects are lost, or images are temporarily blocked; intrusion detection, monitoring of new unknown objects (such as people, animals, vehicles) in the image; detection of remnants, that is, for monitoring Whether the space has newly added fixed objects for detection; trigger detection, triggering specific programs by detecting image changes, such as intrusion conditions, start recording images, notify security, activate specific switches, and ignore images. The screen of the time; the electronic fence and the warning area are preserved by monitoring the specific space, that is, detecting an unknown event for a specific range, and triggering a specific security procedure; object counting, detecting the movement of the image, calculating the entry and exit personnel or The number of objects; object detection, can lock specific objects in the space for monitoring; object tracking, The image change is detected, the moving object is judged by the continuous image, and the moving track is drawn, and the camera can track the object; the object classification can be through the image color information and the relative object information stored in a database. After the image features are compared, the objects are classified.

The above various applications are based on the image color (Hue) information captured by the present invention as a criterion for judging the factors of light and shadow that may be misjudged, and then determining the degree of matching through matching operations, thereby improving the accuracy of detection.

The step of calculating the matching in the processing method using the image color information of the present invention as described in the third figure.

At the beginning of the step, in step S301, the image is taken by a camera, and after digitizing, the red-green-blue (RGB) information of the specific object in the image is extracted, in this case, the image is originally in red-green-blue. In the degree space, the image information is expressed by red, green and blue colors, converted into a color-saturation-luminance chromaticity space by chromaticity space conversion (step S305) (step S307), and the object color is captured (Hue The information is used as a reference. In step S309, a specific matching operation of the color information between the reference image and the captured image is performed to determine the degree of matching (step S311). The matching operation can apply various conventional equations for calculating matching, and is not described here.

Next, the fourth figure is a chromaticity space diagram of the color-saturation-luminance utilized by the present invention.

The fourth figure shows one of the coordinate forms of the color-saturation-luminance chromaticity space. In this example, the color-saturation-luminance chromaticity space is displayed as a cone of cones expressed by information such as color, color saturation and brightness. Coordinates, where the longitudinal direction is represented by the darkest black to the brightest white, expressed in terms of intensity (intensity) i. Where the circular coordinates of the facet are shown on the right side of the figure, the radius of the circle represents the saturation (Saturation) s, the center of the circle is the lowest saturation of 0, and the part close to the circumference represents the maximum saturation; and the circumference is from 0 to 360 degrees. Indicates the color (Hue) h, and the color is sequentially composed of red, yellow, green, Cyan, blue, and magenta.

Only the expression patterns of the color-saturation-luminance chromaticity space are listed above, and the actual implementation is not limited by this.

The image information utilized by the present invention is required to convert the original color expressed by the red-green-blue degree space into the color-saturation-luminance chromaticity space, and exclude the saturation and brightness. Variables, image difference recognition is performed only with color (Hue) as a variable. The color values here can be expressed in degrees from 0 degrees to 360 degrees, each angle representing a color.

The conversion method is as follows, and this example is only one of the embodiments, but the actual implementation is not limited thereto. For example, the brightness is expressed as i=1/3 (R+G+B), which is the average value of red, green, and blue; the saturation is represented by s=1-(3/(R+G+B))*a, Where a is the minimum of red, green, and blue; color can be h=cos^(-1)((0.5*((RG)+(RB))))/((RG)^2+(RB)* (GB)) ^ (0.5)) expression.

Another conventional technique (refer to U.S. Patent No. 6,665,439) performs a chromaticity space in the following mathematical formula, as shown in equation (1):

Imax=max(r,g,b)

Imin=min(r,g,b)

Brightness i=(imax+imin)/2

When imax=imin

Saturation s=0

When i<=(max_value)/2

Saturation s=(imax-imin)/(imax+imin)*max_value

When i>(max_value)/2

Saturation s=(imax-imin)/(max_value-imax-imin)*max_value

R1=(imax-r)/(imax-imin)

G1=(imax-g)/(imax-imin)

B1=(imax-b)/(imax-imin)

When imax=imin, the color h is not defined.

When r=imax, the color h=((b1-g1)π/3

When g=imax, the color h=((2+rl-b1)π/3

When b=imax, the color h=((4+g1-r1)π/3

Where r, g, b are the original color values of each pixel, imax is the maximum value of r, g, b, and imin is the minimum value of r, g, b, including the brightness i and the saturation s, the range is from 0 To the maximum value, if the recording mode is 8 bits, the maximum value is 255, which is the variable max_value in equation (1). The color is represented by an angle from 0 to 2π.

Using the above equation (1), each pixel (r(x, y), g(x, y), b(x, y)) is converted to h(x, y), s(x, y), respectively. With i(x,y).

The fifth figure is a flowchart of an embodiment of a method for processing image color information according to the present invention. The flow of this embodiment is mainly to extract image color (Hue) information as a reference for image motion detection, and then perform quantization, such as calculation matching, by The degree of matching determines the image change and can be applied to various intelligent video surveillance (IVS) systems.

At the beginning of the processing method step, in step S501, a camera performs shooting, in particular, recording (storing to a memory or a specific temporary storage area) successive images, and then, in step S503, performing a chroma space through the image processing unit. Conversion, as in the above embodiment, in particular, converts the chrominance space of the original image into the color-saturation-luminance chromaticity space applied by the method embodiment.

The image described by the color-saturation-luminance chromaticity space includes parameters such as Hue, Saturation, and Intensity. In step S505, the embodiment only captures specific images. The color (Hue) information of the object, excluding the characteristics of saturation and intensity. The objects in the image can be selected by the user or can be judged by automatically detecting the changed content. Another example is S507, which establishes reference information from the image color information originally under the original conditions as the basis for comparison.

In practical implementation, the continuous image is composed of a plurality of frames, so the so-called chroma space conversion or the captured color information are all based on the frame.

Then, the method embodiment continues to capture the subsequent image (step S509), and also performs the chrominance space conversion in step S511. In the embodiment of the present invention, the chromaticity space at the original image is converted into the color-saturation-luminance. The chromaticity space. Then, in step S513, the color (Hue) information of the specific object in the subsequent image is captured, and a matching operation is performed with the previously set reference information, and the matching degree determines whether the reference information matches the color information of the subsequent image (step S515).

In this judging step, the degree of association is mainly obtained by calculating the matching between the reference image and the color information of the subsequently captured image, and the matching operation can refer to other conventional techniques, and the main purpose is to properly quantize The difference between the reference information and subsequent color information.

The above determining the degree of association by the matching operation can determine whether the matching degree meets a specific judgment standard by setting another threshold value, and the threshold value can be used to adjust the judgment criterion, which can be interpreted as sensitivity. For example, if the matching threshold is low (close to the minimum matching degree), it is easy to judge that the matching is low, indicating that the sensitivity is low; if the threshold is high (close to the maximum matching degree), indicating that the matching is likely to be lower, the sensitivity is higher. .

In the step, if it is determined to be a match, the subsequent image is continuously captured (step S509). However, if the determination is not matched, a delay time may be further introduced as one of the conditions for determining, so as to determine the degree of association between the reference image and the subsequent image. In other words, as in step S517, it is judged whether or not a delay time is exceeded. In the determining step of the delay time, it is mainly to determine the degree of association between the reference image and the subsequent image, and also to determine whether the image mismatching condition maintains a delay time, which is used to avoid the possibility of a false positive event. For example, if a specific allowed person enters during the detection process, this is a temporary event, and the delay time is used as an event to exclude the temporary permission. If the delay (or the image is changed) is maintained for a delay time or longer, a message is sent (step S519). Otherwise, if the delay is not maintained, the delay is not maintained, and the temporary event is judged to be excluded. The step of subsequent image (step S509).

After the message is sent, the relevant personnel will process it. This is the subsequent processing and will not be described here. After that, the information related to the message will be released, such as releasing the alarm (step S521), then returning to the initial step S501, re-executing the image capturing, chroma space conversion, capturing color information, establishing reference information, and subsequently capturing images and calculations. Match and other steps.

In particular, if the matching operation between the reference image and the subsequent image and the delay time are judged, if it is determined that the degree of association is low, it may indicate that the position and angle of the camera shooting are changed, or the camera is blocked by the foreign object.

Or, it can detect specific objects in a specific space. If an object is lost, the degree of association between the images (reference matching and time delay) will be displayed, including the difference in color information, and it will be maintained for a while. Or use the information that does not match and maintain for a period of time to determine whether additional objects are added to the specific space.

In another embodiment, if the degree of association between the images is low, a trigger message is sent, such as triggering the camera to start recording an image, or performing other programs, such as stopping recording, issuing a specific alarm, and the like. In another embodiment, the number of contexts of the object is determined to be low, and the event of entering and exiting is judged, thereby starting the counting process.

Further, a flow chart of an object classification embodiment using the image color information processing method of the present invention shown in FIG. 6 is also implemented. This example mainly uses the color information in the image as the comparison judgment as described in the fifth figure above, and this example is applied to the object classification.

First, in step S601, image capturing is performed, and then the object color (Hue) information is extracted for a specific object or space in the image (step S603). Next, in step S605, a database is introduced, in which the color features of various image objects are recorded, in particular, the color characteristics of the image converted to the color-saturation-luminance chromaticity space, through the comparison (step S607), the objects in the captured images can be classified (step S609), and the objects are classified into category one (step S611), category two (step S613), and category three (step S615) as shown in the flowchart. , or more categories.

The system for applying the above image processing method can refer to the schematic diagram of the embodiment of the processing system using the image color information according to the seventh embodiment.

As shown in the figure, the preferred embodiment of the system includes at least an image capturing unit 701 capable of capturing a continuous picture, and an analog digital conversion unit 703 electrically connected to the image capturing unit 701. The image captured by the lens is analogous. It will be digitized by the analog digital conversion unit 703 and temporarily stored in the temporary storage memory 705 electrically connected to each other. The temporary storage memory 705 is usually a flash memory or a dynamic random access memory (RAM) disposed at the front end of the system, or a specific block in the hard disk.

The processing core in the system is a processing unit 707 electrically connected to each circuit. As shown, the system includes a temporary storage memory 705, a storage unit 709, a communication unit 713, and an image processing module 715. The processing unit 707 is configured to process the captured continuous image, and extract the reference image and the subsequent image therefrom. The storage unit 709 is configured to store the reference image and the subsequent image captured in a period of time, and may be connected or built into a database 711. According to the above embodiment, the database 711 specifically records information about each image and object, and records Image information of a plurality of object categories, including color (Hue) features captured in the image object after chroma space conversion.

Another communication unit 713 can connect to the remote control room 70 by wired or wireless communication means, and the control room 70 can remotely control the processing system using the image color information, and receive the message transmitted by the system, such as applied to the smart. On the video surveillance system, by monitoring the control of the computer room 70, the information transmitted by the security system can be immediately received.

The image processing module 715 obtains image information from the storage unit 709, and the system performs color space conversion through the chromaticity space conversion unit 71 in the image processing module 715. In the preferred embodiment, the frames in the continuous image are originally The chromaticity space is converted into a color-saturation-luminance chromaticity space.

Through the color feature extraction unit 72, the chromaticity space-converted color (Hue) feature (which can be targeted to a specific object) is extracted, and the characteristics of saturation and brightness therein are excluded.

Through the matching operation unit 73, the matching value between the reference image and the color features of each frame in the subsequent image is calculated, thereby determining whether the reference image matches each subsequent image by the matching operation, and determining the image variation according to the matching degree. Finally, the threshold is compared to determine the change of the object to determine whether to send a message, such as determining that the object is abnormal, that is, issuing an alert message.

Next, the eighth to eleventh drawings show various embodiments to which the present invention is applied, but the actual implementation is not limited thereto.

The lost monitoring embodiment as shown in the eighth figure.

The camera 80 fixedly monitors the specific space 82. In addition to the fixed capture of the specific space, the embodiment can also periodically rotate the dynamic scan of a specific space 82, and thus the space 82 sets the reference image 801 under an initial condition, and the subsequent image is used. By comparing the comparison image 802, by comparing the color information (Hue) information of the reference image 801 and the comparison image 802, it is possible to determine that a lost event occurs in the monitored space. This may trigger a security system.

The ninth diagram shows an embodiment of the legacy monitoring.

The camera 90 also sets a reference direction for the space 92 or dynamic scanning. The reference image 901 is set by the initial environment, and the comparison image 902 is obtained through the subsequent image. After capturing the color information and comparing, the new image can be obtained. Unknown items. This also triggers the security system.

Similarly, as shown in the tenth embodiment, the embodiment in which the detecting lens is moved is performed, wherein the camera 10 first generates the reference image 101 in the original environment, and then compares the different images in the image 102 to find an event in which the camera is moved. The security system may be triggered for further processing.

Comparing the above reference image with the continuously changing contrast image, a continuously changed image can be obtained, and a continuously changing picture can be integrated, and a moving track can be drawn. As shown in FIG. 11, the camera 11 with a fisheye lens can be transmitted. The whole process is monitored and it is judged that there is a continuously moving object, such as the moving object 112 and the moving object 112' as shown in the figure.

In summary, the processing system, the processing method and the image classification method using the image color information of the present invention mainly perform the chromaticity space conversion by capturing the captured image, and then extracting the image without being judged by the ambient light and the bright and dark image. The color (Hue) information is then judged by the matching operation to determine the abnormality, which has a high judgment rate.

However, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Therefore, equivalent structural changes that are made by using the specification and the contents of the present invention are equally included in the present invention. Within the scope, it is combined with Chen Ming.

21‧‧‧ buildings

23‧‧‧ camera

25‧‧‧Wireless base station

27‧‧‧Monitoring Center

701‧‧‧Image capture unit

703‧‧‧ analog digital conversion unit

705‧‧‧ temporary memory

707‧‧‧Processing unit

709‧‧‧ storage unit

711‧‧‧Database

713‧‧‧Communication unit

70‧‧‧Control room

715‧‧‧Image Processing Module

71‧‧‧Chroma space conversion unit

72‧‧‧Color feature extraction unit

73‧‧‧Matching unit

80, 90, 10, 11 ‧ ‧ camera

82,92,‧‧‧ space

801,901,101‧‧‧Reference image

802,902,102‧‧‧ alignment images

112,112’‧‧‧moving objects

The first figure is a method for identifying the feature of an object through a color chart by a conventional technique;

The second figure is a schematic diagram of an application embodiment of a processing system using image color information according to the present invention;

The third figure is a step of calculating matching in the processing method of using image color information according to the present invention;

The fourth figure is a chromaticity space diagram of color-saturation-luminance utilized by the present invention;

The fifth figure is a flow chart of an embodiment of a method for processing image color information according to the present invention;

The sixth figure is a flow chart of an object classification embodiment using the image color information processing method of the present invention; the seventh figure is a schematic diagram of an embodiment of a processing system using image color information according to the present invention; and the eighth figure is an application of the application of the present invention. The ninth embodiment is the third embodiment of the present invention; the eleventh embodiment is the third embodiment of the present invention;

S501. . . Image capture

S503. . . Chroma space conversion

S505. . . Capture object color information

S507. . . Establish reference information

S509. . . Image capture

S511. . . Chroma space conversion

S513. . . Capture object color information

S515. . . Whether it matches

S517. . . Whether the delay time is exceeded

S519. . . Send a message

S521. . . All clear

Claims (22)

A method for processing image color information includes: capturing a continuous image and setting a reference image; performing a chromaticity space conversion of each frame in the reference image, converting the chromaticity space in which the reference image is originally located a color-saturation-luminance chromaticity space; extracting color information of an object in the reference image, excluding saturation and brightness of the object in the color-saturation-luminance chromaticity space of the reference image Information to establish a reference information; capture a subsequent image; the chrominance space conversion is performed on each frame in the subsequent image, and the chromaticity space in which the subsequent image is originally converted is converted into the color-saturation-luminance color a color space; extracting color information of the object in the subsequent image, excluding the saturation and brightness information of the object in the color-saturation-luminance chromaticity space; defining by a matching threshold value The range determines whether the reference information matches the color information of the object in the subsequent image; if it is determined to be a match, the subsequent image is continuously captured; and if the judgment does not match And maintain a delay time, represents a non-false, issuing a message; if the mismatching is not maintaining the delay time is expressed as false, the step of capturing successive images using color image information processing is performed. The method for processing image color information according to the first aspect of the patent application, wherein the reference image or the subsequent image is originally in a chromaticity space The space is a red-green-blue space, and the image information is expressed by red, green, and blue colors. The method for processing image color information as described in claim 1 is to determine whether a match is made by whether the difference between the reference information and the color information falls within the matching threshold. The method for processing image color information according to claim 3, wherein if the judgment does not match, then it is determined whether the degree of the mismatch reaches the matching threshold, and further includes: if the degree of mismatch does not reach the Matching the threshold value continues to capture subsequent images; and if the degree of mismatch has reached or exceeded the matching threshold, the message is sent. For example, in the method for processing image color information according to the first aspect of the patent application, wherein the mismatch is judged, the moving path of the object is drawn according to the mismatch information of the color information of the object in the subsequent image. The method for processing image color information according to the first aspect of the patent application, wherein when the determination is not matched, determining whether to capture the reference image or the subsequent image according to the mismatch information of the color information of the object in the subsequent image A camera is moved. The method for processing image color information according to the first aspect of the patent application, wherein when the determination is not matched, determining whether to capture the reference image or the subsequent image according to the mismatch information of the color information of the object in the subsequent image A camera is shaded. The method for processing image color information according to item 1 of the patent application, wherein when the determination is not matched, according to the object in the subsequent image The mismatch information of the color information detects that the object is missing. For example, in the processing method of using image color information according to the first aspect of the patent application, when the mismatch is judged, it is determined that other objects are added according to the mismatch information of the color information of the object in the subsequent image. The method for processing image color information according to the first aspect of the patent application, wherein when the determination is not matched, the message is a trigger message, and the reference is triggered according to the mismatch information of the color information of the object in the subsequent image. The image or a camera of the subsequent image begins recording the image. For example, in the processing method of using image color information as described in claim 10, in which the matching is judged, the camera is stopped to record an image. For example, in the processing method of using image color information as described in claim 1, wherein the judging is not matched, a counting procedure is performed according to the mismatch information of the color information of the object in the subsequent image. An image classification method using image color information, comprising: capturing an image; performing a chromaticity space conversion, converting the image from the original chromaticity space to a color-saturation-luminance chromaticity space; Taking the color information of an object in the image, excluding the information of the saturation and brightness of the object in the color-saturation-luminance chromaticity space of the image; introducing a database, wherein the database records various images The color information of the object; the color information of the object and the color information of the image stored in the database; and sorting the object in the captured image according to the comparison. The image classification method according to claim 13, wherein the chromaticity space is converted into a chromaticity space in which the original chromaticity space of the image is converted into a color-saturation-luminance. The image classification method according to claim 14, wherein the step of extracting the color information of the object in the image excludes the saturation of the object in the color-saturation-luminance chromaticity space. Brightness information. The image classification method according to claim 14, wherein the original chromaticity space of the image is a red-green-blue space, and the image information is expressed by red, green and blue colors. A processing system for using image color information, comprising: an image capturing unit for capturing a continuous image; a temporary memory, electrically connecting the image capturing unit for temporarily storing the continuous image; and a processing unit The storage unit and the image capturing unit are electrically connected to the continuous image, and a reference image and a subsequent image are extracted from the continuous image; a storage unit is electrically connected to the processing unit for Storing the reference image and the subsequent image captured during a period of time; an image processing module electrically connected to the processing unit, comprising: a chrominance space conversion unit, the chromaticity space in which each frame in the continuous image is originally located Converting to a color-saturation-luminance chromaticity space; a color feature capture unit that extracts the color feature of an object from the image converted by the chromaticity space, excluding the saturation and the brightness And a matching operation unit, performing a matching operation between the object in the reference image and an object in the continuous image, and determining whether the color features match, wherein: if the color feature is determined to be a match according to the color feature of each object, Then continue to capture subsequent images; and if it is judged as a mismatch according to the color characteristics of each object, and maintains a delay time, indicating that the delay is not misjudged; if the matching situation does not maintain the delay time, indicating that the error is judged, the process continues. Take subsequent images to use the steps of image color information processing. The processing system for using image color information as described in claim 17 further includes an analog-to-digital conversion unit electrically connected to the image capturing unit to perform analog-to-digital conversion of the image. The processing system for using image color information as described in claim 17 further includes a database electrically connected to the storage unit for recording image information of a plurality of object categories. The processing system for using image color information as described in claim 17 further includes a communication unit, and a remote control center can control the processing system using the image color information through wireless or wired communication means, and receive the The message passed by the system. A processing system using image color information as described in claim 20, wherein the message is an alert message. The processing system for using image color information according to claim 17, wherein the chromaticity space in which each frame in the continuous image is originally is a red-green-blue space.