CN114862779A - Image quality evaluation method, electronic device, and storage medium - Google Patents
Image quality evaluation method, electronic device, and storage medium Download PDFInfo
- Publication number
- CN114862779A CN114862779A CN202210443339.3A CN202210443339A CN114862779A CN 114862779 A CN114862779 A CN 114862779A CN 202210443339 A CN202210443339 A CN 202210443339A CN 114862779 A CN114862779 A CN 114862779A
- Authority
- CN
- China
- Prior art keywords
- template
- pixel points
- window
- speckle pattern
- pixel
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Geometry (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the application relates to the field of image processing, and discloses an image quality evaluation method, electronic equipment and a storage medium. The method comprises the following steps: carrying out foreground identification on the speckle pattern to obtain a foreground area of the speckle pattern; counting the occupation ratio of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern; and evaluating the quality of the depth map according to the occupation ratio. According to the scheme, when the quality evaluation is carried out on the depth map based on the voidage, only the voidage corresponding to the foreground object is counted, so that the counting of the voidage is more reliable and robust, and the quality evaluation of the depth is reflected more truly.
Description
Technical Field
The embodiment of the application relates to the field of image processing, in particular to an image quality evaluation method, electronic equipment and a storage medium.
Background
The structured light camera is a camera which obtains depth data by emitting an active infrared light source, is widely applied to the field of three-dimensional face recognition, and has wide use scenes, such as: payment scenes, door lock scenes and rail transit scenes, in which a general person is the subject of an image, and the other backgrounds are clear and even exceed the working distance of the camera itself. And whether the quality of the depth map or the camera structure is deformed or not is evaluated, one important index is the voidage of the depth map, and the conventional voidage statistical method comprises the following steps: the denominator is the total number of pixels in the depth map, and the numerator is the number of pixels with a depth value of 0 in the depth map.
However, in actual use, under the condition that the speckle pattern used for generating the depth map has a lot of backgrounds, the depth recovery algorithm cannot recover the high-quality depth map, so that the number of pixel points with a depth value of 0 in the depth map increases, the error of the calculated void ratio increases, the quality of the depth map cannot be truly evaluated, and service misjudgment can be caused.
Disclosure of Invention
An object of the embodiments of the present application is to provide an image quality evaluation method, an electronic device, and a storage medium, which are used for only counting a voidage corresponding to a foreground object when performing quality evaluation on a depth map based on the voidage, so that the statistics of the voidage is more reliable and robust, and further, the deep quality evaluation is reflected more truly.
In order to solve the above technical problem, an embodiment of the present application provides an image quality evaluation method, including:
carrying out foreground identification on the speckle pattern to obtain a foreground area of the speckle pattern;
counting the occupation ratio of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern;
and evaluating the quality of the depth map according to the occupation ratio.
An embodiment of the present application further provides an electronic device, including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the image quality assessment method described above.
Embodiments of the present application also provide a computer-readable storage medium storing a computer program which, when executed by a processor, implements the above-described image quality evaluation method.
Compared with the prior art, the method and the device have the advantages that the foreground area of the speckle pattern is obtained by carrying out foreground identification on the speckle pattern; counting the occupation ratio of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern; and evaluating the quality of the depth map according to the proportion. According to the scheme, the speckle pattern is distinguished from the background, so that when quality evaluation is carried out on the depth map based on the voidage, the quality of the foreground is concerned more, and only the voidage corresponding to the foreground object is counted, so that the statistics of the voidage is more reliable and robust, and the quality evaluation of the depth is reflected more truly.
Drawings
One or more embodiments are illustrated by the corresponding figures in the drawings, which are not meant to be limiting.
Fig. 1 is a flowchart of an image quality evaluation method provided in an embodiment of the present application;
fig. 2 is a flowchart of an image quality evaluation method provided in an embodiment of the present application;
fig. 3 is a schematic structural diagram of an electronic device provided in an embodiment of the present application.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present application clearer, the embodiments of the present application will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that in the examples of the present application, numerous technical details are set forth in order to provide a better understanding of the present application. However, the technical solution claimed in the present application can be implemented without these technical details and various changes and modifications based on the following embodiments. The following embodiments are divided for convenience of description, and should not constitute any limitation to the specific implementation manner of the present application, and the embodiments may be mutually incorporated and referred to without contradiction.
One embodiment of the present application relates to an image quality evaluation method for performing quality evaluation on a depth map, where an execution subject may be a terminal, a device, or a server capable of performing image processing, as shown in fig. 1, and the image quality evaluation method specifically includes the following steps.
Step 101, performing foreground identification on the speckle pattern to obtain a foreground area of the speckle pattern.
In particular, in a scene where a speckle pattern is acquired by shooting a target object with a depth camera, the target object is generally considered to be the subject of the speckle pattern, namely, referred to as a foreground region, and the other backgrounds are generally relatively open, namely, referred to as background regions. After the speckle pattern is obtained, the speckle pattern can be subjected to foreground identification (namely foreground and background segmentation) by adopting technologies such as image identification and the like, so that a foreground area of the speckle pattern is obtained.
The method for foreground recognition may be used for recognition based on factors such as the size, shape, and image reflectivity (expressed as the size and distribution of pixel gray levels) of the target object in the foreground region, and the specific method for foreground recognition is not limited in this embodiment.
And 102, counting the proportion of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern.
Specifically, after the speckle pattern is acquired, the depth camera may perform parallax search on the speckle pattern and a reference image of the camera to acquire a parallax image corresponding to the speckle pattern; and then, aiming at the parallax values on all the pixel points in the parallax map, calculating the corresponding depth values by utilizing the triangulation principle, and further forming a depth map corresponding to the speckle pattern.
After the depth map corresponding to the speckle pattern is obtained, determining a region corresponding to a foreground region of the speckle pattern in the depth map, and counting the total number of pixel points in the region and the number of hole pixel points. The occupation ratio of the void pixel points in the region, namely the void rate, can be calculated based on the total number of the pixel points in the region and the number of the void pixel points.
The void pixel point can be defined as a pixel point with a depth value of 0 or a pixel point with an invalid value.
And 103, evaluating the quality of the depth map according to the proportion.
Specifically, in evaluating the depth map quality, it is considered that the smaller the voidage of the depth map, the higher the corresponding image quality, and the larger the voidage, the lower the corresponding image quality. In the embodiment, the image quality of the whole depth map is evaluated by the void ratio in the depth map region corresponding to the more concerned foreground region, so that the adopted void ratio can be ensured to be more reliable and robust, and further the quality evaluation of the depth in the depth map is reflected more truly.
For example, a ratio threshold may be set, and when the ratio of the hollow pixel points in the region corresponding to the foreground region in the depth map is smaller than the ratio threshold, it is considered that the quality of the depth map is high, the usability of the depth map is high, and the depth information of the target object can be truly reflected; and when the proportion of the hollow pixel points in the region corresponding to the foreground region in the depth map is not less than the proportion threshold value, the quality of the depth map is considered to be low, and the depth information of the target object cannot be truly reflected.
Compared with the related art, the method and the device have the advantages that the foreground area of the speckle pattern is obtained by performing foreground identification on the speckle pattern; counting the occupation ratio of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern; and evaluating the quality of the depth map according to the proportion. According to the scheme, the speckle pattern is distinguished from the background, so that when quality evaluation is carried out on the depth map based on the voidage, the quality of the foreground is concerned more, and only the voidage corresponding to the foreground object is counted, so that the statistics of the voidage is more reliable and robust, and the quality evaluation of the depth is reflected more truly.
Another embodiment of the present invention relates to an image quality evaluation method, as shown in fig. 2, which is an improvement of the steps of the method shown in fig. 1 in that the process of acquiring the foreground region of the speckle pattern is refined. As shown in fig. 2, the above step 101 may include the following sub-steps.
Substep 1011: and carrying out overexposure detection on the speckle pattern, and determining non-overexposure pixel points.
Specifically, the speckle pattern is generated by focusing coherent light emitted from a laser diode in a structured light camera through a lens and dispersing the coherent light into a plurality of random light rays through a diffractive optical element to generate a random dot pattern, and sensing the pattern projected onto the surface of an object by an infrared camera. When the speckle pattern is generated, exposure conditions presented by each pixel point on the generated speckle pattern are different due to different distances between the surface of the object and the structured light camera. The image overexposure generally means that the gray level of a certain area in an image is 255, the area is determined as an overexposure area at the moment, and pixel points in the overexposure area are determined as overexposure pixel points; and determining other areas except the area as non-overexposure areas, and determining pixel points in the non-overexposure areas as non-overexposure pixel points.
In one example, determining non-overexposed pixel points may be accomplished as follows.
The method comprises the following steps: and aiming at the pixel point with the gray value of 255 in the speckle pattern, constructing a second window by taking the pixel point as the center.
Specifically, a two-dimensional window can be constructed by taking a pixel point with a gray scale value of 255 in the speckle pattern as a center, and the two-dimensional window is marked as a second window; the size of the second window can be customized, for example, a 5 × 5 window, and the number of pixel points in the window is 25.
Step two: and when the gray standard deviation of the pixel point in the second window is smaller than a first threshold value, determining the pixel point as an overexposure pixel point, and determining the residual pixel points except the overexposure pixel point in the speckle pattern as non-overexposure pixel points.
Specifically, the gray-scale average value μ, μ ═ of the pixels in the second window is obtained (Σ I) i ) N, wherein I i Is the ith pixel point in the second windowThe value range of i is 1,2, …, and N is the total number of the pixels in the second window. After the gray level average value of the second window is obtained, calculating the gray level standard deviation sigma of the pixel points in the second window according to a standard deviation formula, wherein sigma is { [ sigma (I) i -μ)] 1/2 and/N. And setting a first threshold value for measuring the gray standard deviation boundary value of the second window when the overexposure pixel point is taken as the center point of the second window. When the gray standard difference of the pixel points in the second window is smaller than a first threshold, the gray value distribution of the pixel points in the second window is relatively balanced, and then the central pixel point in the second window is determined as an overexposed pixel point; and when the gray standard deviation of the pixel points in the second window is not smaller than the first threshold, the gray value distribution of the pixel points in the second window is unbalanced, and the central pixel point in the second window is not determined as the overexposed pixel point. And after all over-exposure pixel points in the speckle pattern are determined, determining the residual pixel points in the speckle pattern except the over-exposure pixel points as non-over-exposure pixel points.
For convenience of labeling, an overexposed pixel point in the speckle pattern is marked as 0, and a non-overexposed pixel point in the speckle pattern is marked as 1, so that a labeled image Mask1 is formed.
Substep 1012: and aiming at each non-overexposure pixel point, constructing a first window by taking the non-overexposure pixel point as a center.
Specifically, a two-dimensional window can be constructed by taking each pixel point as a center and marking the two-dimensional window as a first window for each non-overexposed pixel point (namely the pixel point marked as 1 in Mask 1) in the speckle pattern; the size of the first window can be customized, for example, a 9 × 9 window, and the number of pixels in the window is 81.
Substep 1013: performing similarity matching on a pre-constructed template and a first window, and determining non-overexposed pixel points with matching results larger than a similarity threshold value as pixel points in a foreground region;
the template is a preset image block with the reflectivity characteristics of a foreground area.
A speckle projector in which a laser diode emits coherent light, which is focused by a lens and dispersed into a plurality of random light rays by a diffractive optical element, thereby generating a random dot pattern. The infrared camera senses the pattern projected onto the surface of the object and generates a speckle image. Because the energy of each ray is different among the emitted rays, and then the central energy of the rays is the highest when the target object reflects, each scattered patch has a brightest speckle point pixel, and then the speckle brightness takes the brightest point as the center to obey a two-dimensional normal distribution (two-dimensional Gaussian distribution). And due to the different reflectivity of the surface materials of the object, the object can take three forms: one is that local areas of the speckle pattern retain the characteristic of high brightness, i.e. high reflectivity, such as a human body; the second is that the local area of the speckle image presents darker characteristics, at this time, the contrast ratio is reduced, namely, the reflectivity is low, such as a human face skin area; the third is the background area, where the energy decays quadratically with the source distance, so there is no speckle feature available.
According to the difference of the reflectivity of the foreground area and the reflectivity of the background area to the speckle light, an image block meeting the reflectivity characteristics of the foreground area can be constructed in advance to serve as a template, the template is used for carrying out sliding matching on the non-overexposed area of the speckle pattern, and therefore whether the area which is scratched in the non-overexposed area is the foreground area or not can be determined.
Specifically, the window size of the pre-constructed template may be the same as the first window size, such as a 9 × 9 window, where the number of pixels in the window is 81. And then carrying out similarity matching on the template and each first window, and when the similarity is higher, namely the matching result is greater than a preset similarity threshold, determining a non-overexposed pixel point positioned in the center of the first window as a pixel point in the foreground region. After all the pixel points belonging to the foreground area in the speckle pattern are determined, the foreground area in the speckle pattern can be determined.
In one example, when the pre-constructed template is similarly matched with the first window, the method may include:
and calculating the normalized square difference between the template and the first window by adopting the following formula, and taking the obtained normalized square difference as a matching result of the similarity matching:
wherein, R (x, y) is the normalized square error, (x, y) is the coordinates of non-overexposed pixels in the speckle pattern, (x ', y') is the coordinates of pixels in the template, K (x ', y') is the gray value of pixels at (x ', y'), and I (x + x ', y + y') is the gray value of pixels at (x + x ', y + y').
In general, when performing similar matching, the coordinate position of the non-overexposed pixel should be aligned with the coordinate position of the center pixel point of the template.
In one example, the process of constructing a template may include:
constructing an image block which takes a central pixel point as a center and has a gray value obeying two-dimensional Gaussian distribution as a template; and setting a standard deviation in a two-dimensional Gaussian function corresponding to the two-dimensional Gaussian distribution so that the template has the reflectivity characteristics of the foreground area.
Specifically, according to the refractive index characteristics of the speckle light on the target object in the foreground region analyzed in the foregoing, when the target object reflects, the ray center energy is highest, each scattered patch has a brightest speckle point pixel, and then the speckle brightness obeys two-dimensional normal distribution (two-dimensional gaussian distribution) with the brightest speckle point as the center, so that the template constructed by simulating the characteristics in this embodiment is also an image block with the center pixel point as the center and the gray value obeying the two-dimensional gaussian distribution. This allows a better match calculation between the template and the first window above. In order to make the template have the reflectivity characteristics of the foreground region, the present embodiment adjusts the standard deviation in the two-dimensional gaussian function corresponding to the two-dimensional gaussian distribution to achieve the purpose.
The specific formula of the two-dimensional gaussian function is given as follows:
wherein, σ is standard deviation, and (x, y) corresponds to the coordinates of the pixel points in the template.
According to the previously analyzed refractive index characteristics of the speckle light on the target object in the foreground region, the refractive index characteristics projected on different target objects (human body, human skin) are different, and further distinction needs to be performed according to the shot target object. Therefore, when a template having a reflectance characteristic of the foreground region is provided, the number of templates may be plural, and the reflectance characteristics of the foreground region of each template may be different. The different reflectivity characteristics can be achieved by adjusting the standard deviation in the two-dimensional gaussian function used by the template.
In one example, the templates may include a first template, the standard deviation of the two-dimensional gaussian function corresponding to the first template is smaller than the first standard deviation, and the first template has a high reflectivity characteristic of the foreground region; the similarity threshold corresponding to the first template is recorded as a high reflectivity threshold.
On this basis, the processing of performing similarity matching on the pre-constructed template and the first window and determining the non-overexposed pixel point of which the matching result is greater than the similarity threshold as the pixel point in the foreground region may include:
and performing similarity matching on the pre-constructed first template and the first window, and determining non-overexposed pixel points of which the matching results are greater than the high-reflectivity threshold value as pixel points in the foreground region.
In particular, in the foregoing, due to the different reflectivity of the surface material of the object, three forms are presented, one of which is that the local area of the speckle pattern retains the characteristic of high brightness, i.e. high reflectivity, such as the human body. The standard deviation of the two-dimensional gaussian function corresponding to the template with high reflectivity is usually small, that is, the central coefficient of the template is large, and the surrounding coefficients are small. Therefore, a template can be constructed, so that the standard deviation in the two-dimensional gaussian function corresponding to the template is smaller than the first standard deviation (the value of the first standard deviation is smaller), and the template can have the high-reflectivity characteristic of the foreground region. In this embodiment, the template having the high reflectance characteristic of the foreground region is referred to as a first template. Meanwhile, the similarity threshold used corresponding to the first template is recorded as a high reflectance threshold.
Correspondingly, when the similarity matching is carried out, the similarity matching can be carried out on the first template and the first window, and the non-overexposed pixel point located at the central position in the first window with the matching result larger than the high-reflectivity threshold value is determined as the pixel point in the foreground region.
For convenience of annotation, the non-overexposed pixel point of the Mask1 whose matching result determined by matching with the first template is greater than the high reflectivity threshold may be further marked as 2, so as to form an annotated image Mask 2.
In another example, the templates may include a second template, the standard deviation of the two-dimensional gaussian function corresponding to the second template is greater than a second standard deviation, and the second template has a low-reflectivity characteristic of the foreground region; the similarity threshold corresponding to the second template is noted as the low reflectance threshold.
On this basis, the processing of performing similarity matching on the pre-constructed template and the first window and determining the non-overexposed pixel point of which the matching result is greater than the similarity threshold as the pixel point in the foreground region may include:
and performing similarity matching on the pre-constructed second template and the first window, and determining non-overexposed pixel points with matching results larger than a low reflectivity threshold value as pixel points in the foreground area.
Specifically, in the foregoing, due to the different reflectivity of the surface material of the object, three forms are presented, wherein the second form is that the local area of the speckle pattern presents a darker characteristic, and the contrast ratio is reduced, i.e. the reflectivity is low, such as the area of the human face skin. The standard deviation of the two-dimensional gaussian function corresponding to the template with low reflectivity is usually large, that is, the difference between the central coefficient of the template and the surrounding coefficients is not large. Therefore, a template can be constructed, so that the standard deviation in the two-dimensional Gaussian function corresponding to the template is larger than a second standard deviation (the value of the second standard deviation is larger than the first standard deviation), and the template can have the low-reflectivity characteristic of the foreground area. In this embodiment, the template having the low-reflectance feature of the foreground region is referred to as a second template. Meanwhile, the similarity threshold used corresponding to the second template is recorded as the low reflectance threshold.
Correspondingly, when the similarity matching is carried out, the similarity matching can be carried out on the second template and the first window, and the non-overexposed pixel point located at the central position in the first window with the matching result larger than the low reflectivity threshold value is determined as the pixel point in the foreground region.
For convenience of annotation, the non-overexposed pixel point of the Mask1 whose matching result determined by matching with the second template is greater than the low reflectivity threshold may be further marked as 3, so as to form an annotated image Mask 3.
Finally, the pixels in the foreground region that is finally determined may include the pixel labeled as 2 in the Mask2 and the pixel labeled as 3 in the Mask 3.
Compared with the related technology, the embodiment determines the non-overexposure pixel points by carrying out overexposure detection on the speckle pattern; aiming at each non-overexposure pixel point, constructing a first window by taking the non-overexposure pixel point as a center; performing similarity matching on a pre-constructed template and a first window, and determining non-overexposed pixel points with matching results larger than a similarity threshold value as pixel points in a foreground region; the template is a preset image block with the reflectivity characteristics of the foreground area. The method can quickly identify the pixel points belonging to the foreground region in the speckle pattern so as to conveniently realize that only the voidage corresponding to the foreground object is counted when the quality evaluation is carried out on the depth map based on the voidage, so that the statistics of the voidage is more reliable and robust, and the final target of the quality evaluation of the depth is reflected more truly.
The embodiment of the present application further relates to an electronic device, as shown in fig. 3, including: at least one processor 201; and a memory 202 communicatively coupled to the at least one processor 201; the memory 202 stores instructions executable by the at least one processor 201, and the instructions are executed by the at least one processor 201, so that the at least one processor 201 can execute the image quality evaluation method in the above embodiments.
Where the memory and processor are connected by a bus, the bus may comprise any number of interconnected buses and bridges, the buses connecting together one or more of the various circuits of the processor and the memory. The bus may also connect various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. A bus interface provides an interface between the bus and the transceiver. The transceiver may be one element or a plurality of elements, such as a plurality of receivers and transmitters, providing a means for communicating with various other apparatus over a transmission medium. The data processed by the processor is transmitted over a wireless medium via an antenna, which further receives the data and transmits the data to the processor.
The processor is responsible for managing the bus and general processing and may also provide various functions including timing, peripheral interfaces, voltage regulation, power management, and other control functions. And the memory may be used to store data used by the processor in performing operations.
The embodiment of the application relates to a computer readable storage medium which stores a computer program. The computer program realizes the above-described method embodiments when executed by a processor.
That is, as can be understood by those skilled in the art, all or part of the steps in the method for implementing the embodiments described above may be implemented by a program instructing related hardware, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be understood by those of ordinary skill in the art that the foregoing embodiments are specific examples for carrying out the present application, and that various changes in form and details may be made therein without departing from the spirit and scope of the present application in practice.
Claims (10)
1. An image quality evaluation method, characterized by comprising:
carrying out foreground identification on the speckle pattern to obtain a foreground area of the speckle pattern;
counting the occupation ratio of hollow pixel points in the region corresponding to the foreground region in the depth map corresponding to the speckle pattern;
and evaluating the quality of the depth map according to the occupation ratio.
2. The image quality evaluation method according to claim 1, wherein the foreground recognition of the speckle pattern to obtain a foreground region of the speckle pattern includes:
carrying out overexposure detection on the speckle pattern, and determining non-overexposure pixel points;
aiming at each non-overexposure pixel point, constructing a first window by taking the non-overexposure pixel point as a center;
performing similarity matching on a pre-constructed template and the first window, and determining the non-overexposed pixel points with matching results larger than a similarity threshold value as pixel points in the foreground area;
the template is a preset image block with the reflectivity characteristics of a foreground area.
3. The image quality evaluation method according to claim 2, wherein the detecting overexposure of the speckle pattern and determining non-overexposed pixel points comprises:
aiming at a pixel point with the gray value of 255 in the speckle pattern, a second window is constructed by taking the pixel point as the center;
and when the gray standard deviation of the pixel points in the second window is smaller than a first threshold value, determining the pixel points as overexposed pixel points, and determining the residual pixel points in the speckle pattern except the overexposed pixel points as non-overexposed pixel points.
4. The image quality evaluation method according to claim 2, wherein the similarity matching of the pre-constructed template with the first window comprises:
calculating the normalized square difference between the template and the first window by adopting the following formula, and taking the obtained normalized square difference as the matching result of the similarity matching:
wherein, R (x, y) is the normalized square difference, (x, y) is the coordinate of the non-overexposed pixel in the speckle pattern, (x ', y') is the coordinate of the pixel in the template, K (x ', y') is the gray value of the pixel at (x ', y'), and I (x + x ', y + y') is the gray value of the pixel at (x + x ', y + y').
5. The image quality evaluation method according to claim 2, wherein the process of constructing the template includes:
constructing an image block which takes a central pixel point as a center and has a gray value obeying two-dimensional Gaussian distribution as the template;
and setting a standard deviation in a two-dimensional Gaussian function corresponding to the two-dimensional Gaussian distribution so that the template has the reflectivity characteristics of the foreground area.
6. The image quality evaluation method according to claim 5, wherein the template comprises a first template, the standard deviation of the two-dimensional Gaussian function corresponding to the first template is smaller than a first standard deviation, and the first template has a high-reflectivity feature of the foreground region; recording the similarity threshold corresponding to the first template as a high reflectivity threshold;
the performing similarity matching between the pre-constructed template and the first window, and determining the non-overexposed pixel points with matching results larger than a similarity threshold as the pixel points in the foreground region, includes:
and performing similarity matching on the first template and the first window which are constructed in advance, and determining the non-overexposed pixel points with matching results larger than the high reflectivity threshold value as the pixel points in the foreground area.
7. The image quality evaluation method according to claim 5 or 6, wherein the template comprises a second template, the standard deviation of the two-dimensional Gaussian function corresponding to the second template is greater than a second standard deviation, and the second template has a low-reflectivity characteristic of the foreground region; the similarity threshold corresponding to the second template is recorded as a low reflectivity threshold;
the performing similarity matching between the pre-constructed template and the first window, and determining the non-overexposed pixel points with matching results larger than a similarity threshold as the pixel points in the foreground region, includes:
and performing similarity matching on the pre-constructed second template and the first window, and determining the non-overexposed pixel points with matching results larger than the low reflectivity threshold value as the pixel points in the foreground area.
8. The image quality evaluation method according to claim 1, wherein the void pixel is a pixel having a depth value of 0 or a pixel having an invalid value.
9. An electronic device, comprising:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the image quality assessment method of any one of claims 1 to 8.
10. A computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the image quality evaluation method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210443339.3A CN114862779A (en) | 2022-04-25 | 2022-04-25 | Image quality evaluation method, electronic device, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210443339.3A CN114862779A (en) | 2022-04-25 | 2022-04-25 | Image quality evaluation method, electronic device, and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114862779A true CN114862779A (en) | 2022-08-05 |
Family
ID=82633412
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210443339.3A Pending CN114862779A (en) | 2022-04-25 | 2022-04-25 | Image quality evaluation method, electronic device, and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114862779A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115049658A (en) * | 2022-08-15 | 2022-09-13 | 合肥的卢深视科技有限公司 | RGB-D camera quality detection method, electronic device and storage medium |
| CN115423808A (en) * | 2022-11-04 | 2022-12-02 | 合肥的卢深视科技有限公司 | Quality detection method for speckle projector, electronic device, and storage medium |
-
2022
- 2022-04-25 CN CN202210443339.3A patent/CN114862779A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115049658A (en) * | 2022-08-15 | 2022-09-13 | 合肥的卢深视科技有限公司 | RGB-D camera quality detection method, electronic device and storage medium |
| CN115423808A (en) * | 2022-11-04 | 2022-12-02 | 合肥的卢深视科技有限公司 | Quality detection method for speckle projector, electronic device, and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114862779A (en) | Image quality evaluation method, electronic device, and storage medium | |
| CN111353512B (en) | Obstacle classification method, obstacle classification device, storage medium and computer equipment | |
| US9767371B2 (en) | Systems and methods for identifying traffic control devices and testing the retroreflectivity of the same | |
| US20070189584A1 (en) | Specific expression face detection method, and imaging control method, apparatus and program | |
| US11615547B2 (en) | Light field image rendering method and system for creating see-through effects | |
| US20130107010A1 (en) | Surface segmentation from rgb and depth images | |
| JP2008009769A (en) | Face authentication system and face authentication method | |
| CN110689577B (en) | Active rigid body pose positioning method in single-camera environment and related equipment | |
| US9299006B2 (en) | Authentication device and method | |
| CN110634116B (en) | Facial image scoring method and camera | |
| US20120062749A1 (en) | Human body identification method using range image camera and human body identification apparatus | |
| CN110059579B (en) | Method and apparatus for in vivo testing, electronic device, and storage medium | |
| CN111598065A (en) | Depth image acquisition method, living body identification method, apparatus, circuit, and medium | |
| CN113449704A (en) | Face recognition model training method and device, electronic equipment and storage medium | |
| CN111739099A (en) | Falling prevention method and device and electronic equipment | |
| CN114913175B (en) | Speckle image quality evaluation method, electronic device, and storage medium | |
| CN113436735A (en) | Body weight index prediction method, device and storage medium based on face structure measurement | |
| US20230033352A1 (en) | Time-of-flight object detection circuitry and time-of-flight object detection method | |
| CN114998980A (en) | Iris detection method and device, electronic equipment and storage medium | |
| US20240144575A1 (en) | 3d image sensing device with 3d image processing function and 3d image processing method applied thereto | |
| CN115761458B (en) | Indoor and outdoor environment judging method, electronic equipment and storage medium | |
| WO2023188919A1 (en) | Screw configuration estimation device, screw configuration estimation method, and computer program | |
| US20240027188A1 (en) | 8bit conversion | |
| CN114783041A (en) | Target object recognition method, electronic device, and computer-readable storage medium | |
| CN114926519A (en) | Depth recovery method and device, electronic equipment and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |