TWI838086B - Image analyzation method and image analyzation device - Google Patents

Abstract

An image analyzation method and an image analyzation device are disclosed. The method includes: capturing an external image by a camera lens; analyzing the external image by using an image analyzation technology to determine a target region in the external image; generating a color prompt message according to a color distribution status of the target region, wherein the color prompt message provides statistics information related to the color distribution status; and presenting the external image and the color prompt message at the same time.

Description

Image analysis method and image analysis device

The present invention relates to an image analysis technology, and in particular to an image analysis method and an image analysis device.

Traditionally, to test the imaging quality of a camera lens, you need to first use the camera lens to capture external images, and then analyze the external images in the background to evaluate the imaging quality of the camera lens based on the image analysis results. However, such image analysis technology cannot allow users to obtain the imaging quality of the camera lens at the moment of capturing external images, and it is too late to perform lens calibration in the background after completing image capture.

The present invention provides an image analysis method and an image analysis device, which can improve the efficiency of image analysis.

The embodiment of the present invention provides an image analysis method, which includes: capturing an external image through a camera lens; analyzing the external image using image analysis technology to determine a target area in the external image; generating a color prompt message according to the color distribution of the target area, wherein the color prompt message provides statistical information related to the color distribution; and simultaneously presenting the external image and the color prompt message through a display.

The embodiment of the present invention further provides an image analysis device, which includes a camera lens, a display, a storage circuit and a processor. The processor is coupled to the camera lens, the display and the storage circuit. The processor is used to: capture an external image through the camera lens; analyze the external image using image analysis technology to determine a target area in the external image; generate a color prompt message according to the color distribution of the target area, wherein the color prompt message provides statistical information related to the color distribution; and simultaneously present the external image and the color prompt message through the display.

Based on the above, after capturing an external image through a camera lens, the external image can be analyzed to determine a target area in the external image. According to the color distribution of the target area, a color prompt message can be generated to provide statistical information related to the color distribution. Thereafter, the external image and the color prompt message can be simultaneously presented through a display. In this way, the efficiency of real-time image analysis of external images can be improved.

10: Image analysis device

11: Photography lens

12: Display

13: Storage circuit

14: Processor

21~27,31~34,41~43: Blocks

51: Image

501~503: Target area

511~513: Color prompt message

514: Scene prompt message

S601~S604: Steps

FIG1 is a schematic diagram of an image analysis device according to an embodiment of the present invention.

FIG2 is a schematic diagram of generating a color prompt message according to an embodiment of the present invention.

FIG3 is a schematic diagram of generating a color prompt message according to an embodiment of the present invention.

FIG4 is a schematic diagram of generating a scene prompt message according to an embodiment of the present invention.

Figures 5A to 5C are schematic diagrams showing external images, color prompt messages, and scene prompt messages according to an embodiment of the present invention.

FIG6 is a flow chart of an image analysis method according to an embodiment of the present invention.

FIG1 is a schematic diagram of an image analysis device according to an embodiment of the present invention.

Referring to FIG. 1 , the image analysis device 10 may include a camera, a smart phone, a tablet computer, a head-mounted display, a laptop computer, or other electronic devices with image capture and display functions. The present invention does not limit the type of the image analysis device 10.

The image analysis device 10 includes a camera lens 11, a display 12, a storage circuit 13 and a processor 14. The camera lens 11 is used to capture external images. The external image may include or reflect an image of the external environment captured by the camera lens 11. For example, the camera lens 11 may include an image capture module. The image capture module may include a lens and a photosensitive element to support the image capture function. The present invention does not limit the type of the camera lens 11.

In one embodiment, the camera lens 11 is disposed in the image analysis device 10. In one embodiment, the camera lens 11 is an external camera device and is coupled to the image analysis device 10.

The display 12 is used to display images. For example, the display 12 can be used to display the external image in real time. For example, the display 12 may include a plasma display (Plasma Display), a liquid crystal display (liquid-crystal display, LCD), a thin film transistor liquid crystal display (Thin film transistor liquid crystal display, TFT-LCD), an organic light-emitting diode (Organic Light-Emitting Diode, OLED) and a light-emitting diode display (LED display), etc., and the type of the display 12 is not limited thereto.

The storage circuit 13 can be used to store data. For example, the storage circuit 13 can store the external image. For example, the storage circuit 13 may include a non-volatile storage circuit and a volatile storage circuit. For example, the non-volatile storage circuit may include a read-only memory (ROM), a flash memory, a traditional hard disk drive (HDD), or a similar non-volatile storage medium. For example, the volatile storage circuit may include a dynamic random access memory (DRAM) or a similar volatile storage medium.

The processor 14 is coupled to the camera lens 11, the display 12 and the storage circuit 13. The processor 14 can be used to be responsible for the overall or partial operation of the image analysis device 10. For example, the processor 14 may include a central processing unit (CPU), a graphics processing unit (GPU) or other programmable general-purpose or special-purpose microprocessors, digital signal processors (DSP), programmable controllers, application-specific integrated circuits (ASIC), programmable logic devices (PLD) or other similar devices or combinations of these devices.

In one embodiment, the image analysis device 10 may also include various input/output (I/O) devices and/or peripheral devices such as a network interface card, a mouse, a keyboard, a touch pad, a touch panel, a joystick, a remote control, a battery, a power management circuit, a microphone and/or a speaker, which are not described one by one here.

In one embodiment, the processor 14 can capture external images via the camera lens 11. For example, the processor 14 can instruct the camera lens 11 to capture images of the external environment to obtain the external image. The processor 14 can temporarily store the external image in the storage circuit 13.

In one embodiment, the processor 14 may use image analysis technology to analyze the external image to determine one or more specific areas (also referred to as target areas) in the external image. For example, the processor 14 may obtain pixel values corresponding to at least some pixel positions in the external image. Based on these pixel values, the processor 14 may set the image area in the external image that meets specific conditions as the target area.

In one embodiment, the processor 14 can detect specific objects in the external image through image analysis technology. For example, the specific objects may include various preset types or objects with preset shapes, which are not limited by the present invention. Based on the detection result of the specific object, the processor 14 can determine the target area in the external image. For example, after detecting the specific object in the external image, the processor 14 can set the target area according to the image range occupied by the specific object in the external image or the outline of the specific object, so that the determined target area can cover the specific object. In addition, the shape of the target area can be rectangular, circular or irregular, etc., which is not limited by the present invention.

In one embodiment, the processor 14 may use a trained deep learning model or other image recognition model to perform image recognition on the external image to detect specific objects in the external image. For example, the processor 14 may use a large amount of training data (e.g., training images) to train the image recognition model to perform image recognition and verify the image recognition results of the image recognition model based on corresponding verification data. Based on the verification results, the internal decision parameters (e.g., weight values) of the image recognition model may be updated to improve the detection or recognition accuracy of the image recognition model for the specific object.

In one embodiment, after determining the target area, the processor 14 may generate a prompt message (also referred to as a color prompt message) according to the color distribution of the target area. In particular, the color prompt message may be used to provide statistical information related to the color distribution of the target area. Then, the processor 14 may simultaneously present the external image and the color prompt message via the display 12.

In one embodiment, the specific object may include one or more types of objects. In one embodiment, the specific object may include a first type of object and a second type of object. In particular, the first type of object carries a pattern for performing color deviation analysis (also called a color test pattern or a color detection pattern), while the second type of object does not carry the color test pattern. The color test pattern may carry blocks of different colors to detect the imaging quality of the camera lens 11.

In one embodiment, in response to a specific object in the external image being a first-class object, the processor 14 may generate a corresponding color prompt message (also referred to as a first prompt message) according to the color distribution of the target area covering the specific object. In particular, the first prompt message may be used to provide information related to the color deviation of the color test pattern in the first-class object.

In one embodiment, in response to a specific object in the external image being a first-class object, the processor 14 may perform color detection on a target area covering the specific object to obtain information related to the color deviation of the color test pattern in the first-class object. For example, the processor 14 may obtain the (actual) pixel value corresponding to each pixel position in the color test pattern in the target area. The processor 14 may compare these pixel values with the standard pixel value corresponding to each pixel position in the color test pattern. The processor 14 may obtain the difference between the (actual) pixel value of each pixel position in the color test pattern and the corresponding standard pixel value based on the comparison result. The processor 14 may obtain the color deviation amount of each pixel position in the color test pattern based on the difference. Then, the processor 14 can generate a first prompt message based on the color deviation amount.

FIG2 is a schematic diagram of generating a color prompt message according to an embodiment of the present invention.

Referring to FIG. 2 , in block 21 , the processor 14 may activate the color detection module. In block 22 , the processor 14 may detect the color detection pattern (i.e., the first type of object) in the external image through the color detection module. For example, the color detection module may be designed, trained, and used in conjunction with the aforementioned deep learning model to detect the color detection pattern (i.e., the first type of object) in the external image.

After detecting the color detection pattern in the external image, in block 23, the processor 14 may determine whether the color detection pattern is a known pattern. For example, the processor 14 may perform edge comparison or color block comparison on the color detection pattern and one or more predicted detection patterns in the database. If the comparison result shows that the color detection pattern matches a preset detection pattern in the database (also referred to as the first preset detection pattern), the processor 14 may determine that the color detection pattern is a known pattern. In response to the color detection pattern being a known pattern, in block 24, the processor 14 may apply the first preset detection pattern in the database as a detection sample (also referred to as a preset detection sample).

On the other hand, if the comparison result shows that the color detection pattern does not match any of the preset detection patterns in the database, the processor 14 may determine that the color detection pattern is not a known pattern. In response to the color detection pattern not being a known pattern, in block 25, the processor 14 may apply a similar detection pattern in the database that is closest to the color test pattern as a detection sample (also referred to as a similar detection sample). For example, the similar detection pattern may include a color test pattern with different color blocks commonly seen on the Internet.

Next, in block 26, the processor 14 may determine the target area in the external image according to the detected color test pattern (or the first type of object). For example, the processor 14 may set the image area in the external image that covers the color detection pattern (or the first type of object) as the target area. In block 27, the processor 14 may compare the pixel values corresponding to each pixel position in the color test pattern in the target area with the pixel values corresponding to the same pixel position in the detection sample (i.e., the standard pixel values) and generate a color prompt message (i.e., the first prompt message) according to the comparison result. The first prompt message may include information about the color deviation related to the color test pattern in the external image to reflect the color deviation between the (actual) pixel values of each pixel position in the color test pattern and the corresponding standard pixel values. Based on the information of the color deviation, the user and/or the processor 14 can know the image quality of the current camera lens 11.

In one embodiment, the information related to the color deviation of the color test pattern (e.g., the color deviation amount) provided by the first prompt message can be used to adjust at least some of the setting parameters of the camera lens 11. For example, after adjusting at least some of the setting parameters of the camera lens 11, the color deviation amount corresponding to the color test pattern in the recaptured external image can be reduced, thereby improving the image quality of the external image captured by the camera lens 11.

Please return to FIG. 1. In one embodiment, in response to a specific object in the external image being a second type of object, the processor 14 may generate a corresponding color prompt message (also referred to as a second prompt message) according to the color distribution of the target area covering the specific object. In particular, the second prompt message may be used to provide information related to the color distribution of the second type of object.

In one embodiment, in response to a specific object in an external image being a second-class object, the processor 14 may perform color detection on a target area covering the specific object to obtain information related to the color distribution of the second-class object. For example, the processor 14 may obtain the pixel value corresponding to each pixel position in the object image in the target area. The processor 14 may calculate the numerical distribution of these pixel values. For example, this numerical distribution may reflect the statistical information of the pixel value corresponding to each pixel position in the object image. For example, this statistical information may reflect the number of pixel positions with pixel values of 0 to 255 in the object image, etc., and this statistical information may be designed and adjusted according to practical needs. Then, the processor 14 may generate a second prompt message based on this numerical distribution.

FIG3 is a schematic diagram of generating a color prompt message according to an embodiment of the present invention.

Referring to FIG. 3 , in block 31 , the processor 14 may activate the object detection module. In block 32 , the processor 14 may detect the second type of object (also referred to as the target object) in the external image through the object detection module. For example, the object detection module may be designed, trained and used in conjunction with the aforementioned deep learning model to detect the target object in the external image.

After detecting the target object in the external image, in block 33, the processor 14 may determine the target area in the external image according to the detected target object. For example, the processor 14 may set the image area in the external image that covers the target object as the target area. Then, in block 34, the processor 14 may perform color detection on the target area that covers the target object and generate a color prompt message (i.e., a second prompt message) according to the detection result. For example, the second prompt message may reflect the numerical distribution state of the pixel values of at least part of the pixel positions in the target object in the target area. This numerical distribution state may also reflect the imaging quality of the current camera lens 11. In one embodiment, this numerical distribution state can also be used to adjust at least some of the setting parameters of the camera lens 11 to improve the imaging quality of the camera lens 11.

Taking a taxi in an external image as an example of a target object, under normal circumstances, the body of the taxi is mostly yellow. Therefore, if the second prompt message reflects that the color distribution in the target area of the image containing the taxi is mostly yellow, it means that the imaging quality of the current camera lens 11 is reasonable. However, if the second prompt message reflects that the color distribution in the target area of the image containing the taxi has only a small amount of yellow, it means that the imaging quality of the current camera lens 11 is unreasonable and needs to be corrected.

Please go back to Figure 1. In one embodiment, the processor 14 can also use image analysis technology to analyze the external image to detect the scene type corresponding to the external image. The scene type can reflect the scene type of the environment corresponding to the external image captured by the current camera lens 11, such as indoors, outdoors, in the sun, cloudy, rainy, daytime and/or nighttime, etc., and the scene type is not limited thereto. For example, the processor 14 can determine the scene type corresponding to the external image based on the overall color tone (such as color temperature) of the external image or other reference basis.

In one embodiment, the processor 14 may generate a prompt message (also referred to as a scene prompt message) according to the detected scene type. This scene prompt message may reflect the environmental state corresponding to the external image captured by the current camera lens 11. Then, the processor 14 may simultaneously present the external image, the color prompt message, and the scene prompt message through the display 12.

FIG4 is a schematic diagram of generating a scene prompt message according to an embodiment of the present invention.

Please refer to FIG. 4 . In block 41 , the processor 14 may activate the scene detection module. In block 42 , the processor 14 may detect the scene type corresponding to the external image through the scene detection module. For example, the scene detection module may be designed, trained and used in conjunction with the aforementioned deep learning model to detect the scene type corresponding to the external image.

After detecting the scene type corresponding to the external image, in block 43, the processor 14 may generate a scene prompt message according to the detected scene type. For example, the scene prompt message may reflect that the scene type corresponding to the external image is indoors, outdoors, in the sun, cloudy, rainy, daytime and/or nighttime, etc.

In one embodiment, the overall color tone of the captured external image may be different under different scene types (e.g., different weather conditions, indoors or outdoors, etc.), which may cause errors in the color deviation of the detected color test pattern and/or the color distribution of the target object. In one embodiment, the processor 14 can adjust at least part of the parameters of the camera lens 11 according to the scene type corresponding to the currently detected external image, combined with the aforementioned color deviation and/or color distribution information, so as to improve the accuracy of the parameter adjustment of the camera lens 11. In this way, errors in the parameter adjustment of the camera lens 11 due to different environmental types can be avoided.

In one embodiment, the processor 14 may present the color prompt message (i.e., the first prompt message and/or the second prompt message) and/or the scene prompt message in synchronization with the external image in the display interface of the display 12. For example, the processor 14 may instruct the display 12 to float the color prompt message and/or the scene prompt message above the layer of the external image.

Figures 5A to 5C are schematic diagrams showing external images, color prompt messages, and scene prompt messages according to an embodiment of the present invention.

Please refer to FIG. 5A , assuming that the external image includes image 51. The scene type corresponding to image 51 is outdoor, and image 51 includes a first type of object (i.e., a color test pattern) and a second type of object (i.e., a vehicle and a human face).

Please refer to FIG. 5B . After analyzing image 51 through image analysis technology, the first type of objects (i.e., color test patterns) and the second type of objects (i.e., vehicles and human faces) can be detected, and target areas 501 to 503 can be determined. Among them, target area 501 covers the first type of objects (i.e., color test patterns), and target areas 502 and 503 cover the second type of objects (i.e., human faces and vehicles), as shown in FIG. 5B .

Referring to FIG. 5C , after detecting and analyzing the color distribution of the target area, color prompt messages 511-513 may be presented (e.g., floated) above the layer of the image 51 to provide statistical information related to the color distribution of the target area. The color prompt message 511 may provide information related to the color deviation of the color test pattern in the target area 501. For example, the length of the connection between any two points in the color prompt message 511 may reflect the color deviation of a certain color block in the color test pattern. The longer the length of the connection between any two points in the color prompt message 511, the greater the color deviation of the color block corresponding to the connection. Color prompt message 512 may provide information related to the color distribution of the target object in target area 502 (e.g., statistical information related to the color distribution of the face in target area 502). Color prompt message 513 may provide information related to the color distribution of the target object in target area 503 (e.g., statistical information related to the color distribution of the vehicle in target area 503). It should be noted that the statistical information related to the color distribution of the target area provided by color prompt messages 511-513 may be designed and adjusted according to practical needs, and the present invention is not limited thereto.

In one embodiment, after scene detection is performed on image 51, scene prompt message 514 may also be presented in image 51 (e.g., floating above the layer of image 51) according to the scene type corresponding to image 51. For example, scene prompt message 514 may provide scene information of image 51 (e.g., outdoor).

In one embodiment, when the color prompt messages 511-513 (and the scene prompt message 514) are presented in real time, the imaging quality of the camera lens 11 can be evaluated in real time when the image 51 is captured according to the color prompt messages 511-513 (and the scene prompt message 514). This evaluation result can be used to adjust at least part of the setting parameters of the camera lens 11 in real time or at any subsequent time point. In this way, the parameter adjustment efficiency of the camera lens 11 can be effectively improved.

FIG6 is a flow chart of an image analysis method according to an embodiment of the present invention.

Please refer to Figure 6. In step S601, an external image is captured by a camera lens. In step S602, the external image is analyzed using image analysis technology to determine the target area in the external image. In step S603, a color prompt message is generated according to the color distribution of the target area, wherein the color prompt message provides statistical information related to the color distribution. In step S604, the external image and the color prompt message are simultaneously presented through a display.

However, each step in FIG6 has been described in detail above and will not be repeated here. It is worth noting that each step in FIG6 can be implemented as multiple program codes or circuits, which is not limited in this case. In addition, the method of FIG6 can be used in conjunction with the above exemplary embodiments or can be used alone, which is not limited in this case.

In summary, the image analysis method and image analysis device proposed in the embodiment of the present invention can present the relevant color distribution or color deviation information together with the external image captured by the camera lens on the display. In this way, the user can easily adjust and calibrate at least part of the setting parameters of the camera lens according to this information without waiting for the user to return home or use other professional instruments, thereby improving the imaging quality of the camera lens.

Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the relevant technical field may make some changes and modifications within the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of the patent application attached hereto.

S601~S604: Steps

Claims (12)

An image analysis method includes: capturing an external image through a camera lens; using an image recognition model to analyze the external image to detect a specific object from the external image; determining a target area in the external image according to the detected specific object; generating a color prompt message according to the color distribution of the target area, wherein the color prompt message provides statistical information related to the color distribution; and simultaneously presenting the external image and the color prompt message through a display. An image analysis method as described in claim 1, wherein the specific object includes a first type of object and a second type of object, the first type of object has a color test pattern, and the second type of object does not have the color test pattern. The image analysis method as described in claim 2, wherein the step of generating the color prompt message according to the color distribution of the target area includes: in response to the specific object being a first type of object, generating a first prompt message according to the color distribution of the target area, wherein the first prompt message provides information related to the color deviation of the color test pattern. The image analysis method as described in claim 2, wherein the step of generating the color prompt message according to the color distribution of the target area includes: in response to the specific object being a second type of object, generating a second prompt message according to the color distribution of the target area, wherein the second prompt message provides information related to the color distribution of the second type of object. The image analysis method as described in claim 1 further includes: using image analysis technology to analyze the external image to detect the scene type corresponding to the external image; generating a scene prompt message according to the scene type; and simultaneously presenting the external image, the color prompt message and the scene prompt message through the display. The image analysis method as described in claim 1, wherein the step of simultaneously presenting the external image and the color prompt message via the display includes: floating the color prompt message above the layer of the external image. An image analysis device includes: a camera lens; a display; a storage circuit; and a processor coupled to the camera lens, the display and the storage circuit, wherein the processor is used to: capture an external image through the camera lens; use an image recognition model to analyze the external image to detect a specific object from the external image; determine a target area in the external image according to the detected specific object; generate a color prompt message according to the color distribution of the target area, wherein the color prompt message provides statistical information related to the color distribution; and simultaneously present the external image and the color prompt message through the display. An image analysis device as described in claim 7, wherein the specific object includes a first type of object and a second type of object, the first type of object has a color test pattern, and the second type of object does not have the color test pattern. The image analysis device as described in claim 8, wherein the operation of the processor generating the color prompt message according to the color distribution of the target area includes: in response to the specific object being a first type of object, generating a first prompt message according to the color distribution of the target area, wherein the first prompt message provides information related to the color deviation of the color test pattern. The image analysis device as described in claim 8, wherein the operation of the processor generating the color prompt message according to the color distribution of the target area includes: in response to the specific object being a second type of object, generating a second prompt message according to the color distribution of the target area, wherein the second prompt message provides information related to the color distribution of the second type of object. The image analysis device as described in claim 7, wherein the processor is further used to: analyze the external image using image analysis technology to detect the scene type corresponding to the external image; generate a scene prompt message according to the scene type; and simultaneously present the external image, the color prompt message and the scene prompt message through the display. The image analysis device as described in claim 7, wherein the operation of the processor simultaneously presenting the external image and the color prompt message via the display includes: floating the color prompt message above the layer of the external image.

Images