Enhancement method, device, equipment and storage medium for unmanned equipment collected images

Abstract

The embodiment of the present invention discloses an enhancement method, device, equipment and storage medium for images collected by unmanned equipment, and the present solution relates to the field of remote sensing technology. The method comprises: obtaining the original image collected by the unmanned equipment, extracting the parameter information of the original image; when it is identified that the parameter information satisfies the first enhancement processing condition, performing the first enhancement processing, and / or, when it is identified that the parameter information satisfies the second enhancement processing condition, performing the second enhancement processing, and performing contrast enhancement processing to obtain the enhancement processing result of the original image. The present technical solution can solve the problem that the images collected by unmanned equipment during agricultural operations are affected by factors such as weather and environment, resulting in poor image acquisition results and failure to normally reflect the growth status of crops, by performing the first enhancement processing and / or the second enhancement processing at the same time as the contrast enhancement processing, thereby achieving multi-faceted improvements in image quality.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of remote sensing technology, and in particular to a method, device, equipment and storage medium for enhancing images collected by unmanned equipment. Background Art

[0002] Remote sensing technology is a technology that obtains information about the earth's surface through flying equipment. With the development of science and technology, remote sensing technology has been applied in various fields. Among them, agricultural remote sensing drones can be used for farmland monitoring, crop growth monitoring, and agricultural resource management, helping to improve production efficiency and resource utilization efficiency.

[0003] However, the agricultural operating environment has the characteristics of large ground span, changeable scenes and high shooting altitude. The quality of images collected by agricultural remote sensing drones is uneven. In addition, environmental factors such as dark night scene images and unclear foggy images, as well as low local contrast in large-area monotonous scenes, will lead to poor acquisition of single images, which in turn affects the stitching effect of color high-definition images, resulting in low efficiency and poor results of agricultural remote sensing drone operations, as well as greatly compressed operation time limits.

[0004] At present, shooting can only be carried out when the weather conditions are good and the light intensity is sufficient, resulting in the operating efficiency being severely limited by environmental factors, and whether the image quality is qualified still needs to be considered in other aspects. Therefore, how to improve the quality of image data collected by agricultural remote sensing drones is a technical problem that technicians in this field need to solve urgently. Summary of the invention

[0005] The embodiments of the present invention provide an enhancement method, device, equipment and storage medium for images collected by unmanned equipment. By performing a first enhancement process and / or a second enhancement process while performing a contrast enhancement process, low-quality images can be optimized so that images collected by the unmanned equipment in different weather and environments can normally reflect the growth status of crops.

[0006] In a first aspect, an embodiment of the present invention provides a method for enhancing images collected by unmanned equipment, the method comprising:

[0007] Acquire the original image collected by the unmanned equipment, and extract parameter information of the original image;

[0008] When it is identified that the parameter information satisfies the first enhancement processing condition, the first enhancement processing is performed, and / or,

[0009] When it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and the contrast enhancement processing is performed to obtain the enhancement processing result of the original image.

[0010] In a second aspect, an embodiment of the present invention further provides an enhancement device for collecting images by unmanned equipment, the device comprising:

[0011] A parameter information extraction module is used to obtain the original image collected by the unmanned equipment and extract the parameter information of the original image;

[0012] a first enhancement processing module, configured to perform first enhancement processing when it is identified that the parameter information meets a first enhancement processing condition, and / or

[0013] The second enhancement processing module is used to perform the second enhancement processing and the contrast enhancement processing to obtain the enhancement processing result of the original image when it is identified that the parameter information meets the second enhancement processing condition.

[0014] In a third aspect, an embodiment of the present invention further provides an enhancement device for images collected by unmanned equipment, the device comprising: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the enhancement method for images collected by unmanned equipment described in an embodiment of the present invention.

[0015] In a fourth aspect, an embodiment of the present invention further provides a storage medium storing computer executable instructions, which, when executed by a computer processor, are used to execute the method for enhancing images collected by unmanned equipment described in an embodiment of the present invention.

[0016] In an embodiment of the present invention, an original image collected by an unmanned device is obtained, and parameter information of the original image is extracted; when it is identified that the parameter information satisfies a first enhancement processing condition, a first enhancement processing is performed, and / or, when it is identified that the parameter information satisfies a second enhancement processing condition, a second enhancement processing is performed, and contrast enhancement processing is performed to obtain an enhancement processing result of the original image. This solution can solve the problem that the images collected by unmanned equipment during agricultural operations are affected by factors such as weather and environment, resulting in poor image collection results and failure to normally reflect the growth status of crops, by performing the first enhancement processing and / or the second enhancement processing at the same time as the contrast enhancement processing, thereby achieving multi-faceted improvements in image quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of a flow chart of a method for enhancing images collected by unmanned equipment provided in Embodiment 1 of the present invention;

[0018] Figure 2 A schematic diagram of a flow chart of a method for enhancing images collected by unmanned equipment provided in Embodiment 2 of the present invention;

[0019] Figure 3 A schematic diagram of a flow chart of a method for enhancing images collected by unmanned equipment provided in Embodiment 3 of the present invention;

[0020] Figure 4 A schematic diagram of the structure of an enhancement device for collecting images by unmanned equipment provided in Embodiment 4 of the present invention;

[0021] Figure 5 A schematic diagram of the structure of an enhancement device for collecting images by unmanned equipment provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only the part related to the present application but not all the contents are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0024] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0025] The following is a detailed description of the method for enhancing images collected by unmanned equipment provided in the embodiment of the present application through specific embodiments and their application scenarios in conjunction with the accompanying drawings.

[0026] Embodiment 1

[0027] Figure 1 The flowchart of the method for enhancing the image collected by unmanned equipment provided in the first embodiment of the present invention is as follows. Figure 1 As shown, the specific steps include:

[0028] S101, obtaining an original image collected by an unmanned device, and extracting parameter information of the original image.

[0029] First, this application is applicable to the scene of using remote sensing technology to obtain crop images. Specifically, the extraction of parameter information and the first enhancement processing, the second enhancement processing and the contrast enhancement processing can be performed by the chip, and the staff can obtain more accurate and comprehensive crop information based on the more complete and better images after processing.

[0030] Based on the above usage scenarios, it can be understood that the execution subject of the present application can be the chip, which includes but is not limited to CPU, DSP, GPU and NPU. Among them, CPU (Central Processing Unit) is one of the core components in the computer system, also known as a microprocessor or processor, which is responsible for executing computer program instructions, controlling data flow and coordinating the main components of various hardware devices; DSP (Digital Signal Processor) is a microprocessor specially designed to perform digital signal processing tasks, with high-speed computing power, parallel processing capabilities and a special instruction set, which can effectively execute various signal processing algorithms; GPU (Graphics Processing Unit) is a processor specially used for processing graphics and image data, used to accelerate tasks such as graphics rendering, image processing and parallel computing; NPU (Neural Processing Unit) is a processor specially used for neural network calculations, and is a hardware accelerator designed for artificial intelligence and machine learning tasks.

[0031] Unmanned equipment can be remote sensing drones and unmanned vehicles with image acquisition capabilities. Remote sensing drones are unmanned aerial vehicles used to collect and monitor remote sensing data. They can obtain high-resolution and high-precision remote sensing data through autonomous or remotely controlled flight. Remote sensing drones are generally equipped with various sensors and imaging equipment, such as cameras, infrared thermal imagers, and multispectral sensors. These devices can obtain images, spectra, and other related data targeting the surface, atmosphere, and water bodies.

[0032] The original image can be the image data collected by the camera installed on the unmanned equipment. Image data is a data set composed of digital pixels, which is used to represent the visual information in the image. Each pixel contains the value of the color and brightness of a certain position in the image.

[0033] The process of unmanned equipment collecting raw images includes: the optical sensor located behind the camera lens converts the light entering through the lens into an electrical signal; the photosensitive element in the optical sensor detects the intensity and color of the light and converts it into a voltage signal; the analog-to-digital converter converts the continuously changing analog voltage signal into digital form for further processing and storage; the image data is output to the chip through an interface (such as USB, HDMI, and network cable).

[0034] The parameter information may be an image parameter used to identify whether to perform the first enhancement processing and / or the second enhancement processing. The parameter information may be extracted by extracting at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast, and saturation of the original image, or by dividing the original image according to a preset rule to obtain at least two image blocks, and calculating the contrast of each image block.

[0035] S102: When it is identified that the parameter information satisfies a first enhancement processing condition, perform a first enhancement processing, and / or.

[0036] The first enhancement processing condition may be that the unmanned device is in a low-light environment and the brightness of the original image needs to be optimized. The first enhancement processing is dark light enhancement processing. The method of identifying whether the parameter information meets the first enhancement processing condition may be to construct a feature vector composed of at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation, and input the feature vector into a pre-obtained machine learning model. If the machine learning model outputs a first result, it is determined that the first enhancement processing condition is met. If the machine learning model outputs a second result, it is determined that the first enhancement processing condition is not met.

[0037] In this scheme, whether the parameter information satisfies the first enhanced processing condition can be pre-set. When the result is that the first enhanced processing condition is satisfied, the relevant process of the first enhanced processing is performed; when the result is that the first enhanced processing condition is not satisfied, the relevant process of the first enhanced processing is directly skipped to determine whether the second enhanced processing condition is satisfied.

[0038] S103, when it is identified that the parameter information meets the second enhancement processing condition, performing the second enhancement processing and performing the contrast enhancement processing to obtain the enhancement processing result of the original image.

[0039] The second enhancement processing condition may be a situation where the unmanned equipment is in a foggy environment and the original image needs to be defogged. The second enhancement processing condition is the defogging enhancement processing. The second enhancement processing condition identifies whether the parameter information satisfies the second enhancement processing condition. If the variance of the contrast of each image block is less than a set threshold, it is determined that the pre-set second enhancement processing condition is met. If the variance of the contrast of each image block is less than the set threshold, it is determined that the second enhancement processing condition is met. If the variance of the contrast of each image block is not less than the set threshold, it is determined that the second enhancement processing condition is not met.

[0040] Contrast refers to the difference between the brightness values ​​in an image, reflecting the difference between light and dark in different areas of the image. Contrast enhancement processing includes two steps: determining the optimal knee point position through the Auto Knee algorithm and adjusting the contrast curve according to the optimal knee point position.

[0041] The knee point represents the turning point where the brightness value changes from linear to non-linear compression, and is a key parameter for preserving details in the shadow and highlight areas of the image. The Auto Knee algorithm automatically analyzes the brightness distribution and adjusts the knee point accordingly to preserve details in the shadow and highlight areas.

[0042] According to the method of adjusting the contrast curve based on the optimal knee point position, the optimal knee point position can be used as a parameter for adjusting the contrast curve, the brightness range can be divided into two areas, dark area and bright area, the dark area is compressed and the bright area is stretched, so as to enhance the contrast of the image. Among them, the contrast curve defines the mapping relationship between the pixel value of the input image and the pixel value of the output image. By modifying the shape and slope of the contrast curve, the brightness difference of different brightness areas in the image can be changed, so as to enhance or weaken the contrast of the image.

[0043] The enhancement processing result may be an image obtained by subjecting the original image to the first enhancement processing and / or the second enhancement processing and the contrast enhancement processing. The execution order of the three steps of the first enhancement processing, the second enhancement processing and the contrast enhancement processing is not limited. It is understandable that when only the first enhancement processing and the contrast enhancement processing are performed, the original image may be input to the first enhancement processing, and the output result of the first enhancement processing may be input to the contrast enhancement processing. Alternatively, the original image may be input to the contrast enhancement processing, and the output result of the contrast enhancement processing may be input to the first enhancement processing. The processing order of the two may be reversed. Similarly, when only the second enhancement processing and the contrast enhancement processing are performed, or when the first enhancement processing, the second enhancement processing and the contrast enhancement processing are performed, their processing order may also be reversed, and this solution does not specifically limit this.

[0044] In the technical solution, optionally, when it is identified that the parameter information meets the first enhancement processing condition, the first enhancement processing is performed, and / or, when it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and the contrast enhancement processing is performed to obtain the enhancement processing result of the original image, including:

[0045] When it is identified that the parameter information satisfies a first enhancement processing condition, performing a first enhancement processing;

[0046] When the first enhancement processing result satisfies the second enhancement processing condition, performing the second enhancement processing to obtain the second enhancement processing result;

[0047] Performing contrast enhancement processing on the second enhancement processing result to obtain an enhancement processing result of the original image;

[0048] or,

[0049] When it is identified that the parameter information satisfies the second enhancement processing condition, performing the second enhancement processing;

[0050] When the second enhancement processing result satisfies the first enhancement processing condition, performing the first enhancement processing to obtain the first enhancement processing result;

[0051] The first enhancement processing result is subjected to contrast enhancement processing to obtain an enhancement processing result of the original image.

[0052] The advantage of such a setting of the present scheme is that the same enhancement processing result can be obtained by adopting different execution orders for the three steps of the first enhancement processing and / or the second enhancement processing and the contrast enhancement processing. At the same time, by giving priority to the first enhancement processing and / or the second enhancement processing and finally performing the contrast enhancement processing, it is possible to avoid the influence of the contrast enhancement processing on the processing condition judgment of the first enhancement processing and the second enhancement processing, and the processing parameters can be uniformly adjusted during the contrast enhancement processing to avoid the problem of distortion of the image output after enhancement.

[0053] In the example of the present application, the original image collected by the unmanned equipment is obtained, and the parameter information of the original image is extracted; when it is identified that the parameter information meets the first enhancement processing condition, the first enhancement processing is performed, and / or, when it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and contrast enhancement processing is performed to obtain the enhancement processing result of the original image. This technical solution can solve the problem that the images collected by unmanned equipment during agricultural operations are affected by factors such as weather and environment, resulting in poor image acquisition results and failure to normally reflect the growth status of crops, by performing the first enhancement processing and / or the second enhancement processing while performing the contrast enhancement processing, thereby achieving multi-faceted improvements in image quality.

[0054] Embodiment 2

[0055] Figure 2 It is a flow chart of the enhancement method of the image collected by unmanned equipment provided in the second embodiment of the present application. This scheme makes a better improvement on the above embodiment, and the specific improvement is: extracting the parameter information of the original image, including: extracting at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation of the original image; when it is identified that the parameter information meets the first enhancement processing condition, performing the first enhancement processing, including: when at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation meets the first enhancement processing condition, performing dark light enhancement processing.

[0056] like Figure 2 As shown, the specific steps include:

[0057] S201, obtaining original images collected by unmanned equipment.

[0058] S202, extracting at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation of the original image.

[0059] The brightness mean refers to the average brightness value of all pixels in the image. The brightness value can be the grayscale value of a pixel in a grayscale image. To extract the brightness mean, the original image can be converted into a grayscale image, each pixel in the grayscale image is traversed, the brightness value of each pixel is added to obtain a cumulative brightness value, and the cumulative brightness value is divided by the total number of pixels to obtain the brightness mean.

[0060] A histogram is a chart that represents the statistical information of each brightness level in an image. The histogram can show the proportion of the number of pixels at each brightness level in the total number of pixels. A low brightness histogram can be a chart that only retains the statistical information of low brightness levels, showing the proportion of the number of pixels at low brightness levels in the total number of pixels. Low brightness can be the first 10% of the brightness range (0-255), that is, 0-25 or 0-26. The method of extracting the proportion of low brightness histogram can be to convert the original image into a grayscale image, traverse each pixel in the grayscale image, count the total number of pixels and the number of pixels at each low brightness level, and divide the number of pixels at the low brightness level by the total number of pixels to obtain the low brightness histogram proportion.

[0061] The signal-to-noise ratio is an important indicator for measuring image quality, which indicates the relative intensity ratio of the useful signal to the noise in the image. The higher the signal-to-noise ratio, the better the image quality. To extract the signal-to-noise ratio, the original image can be divided into a useful signal area and a noise area, and the average grayscale value or standard deviation of the pixel value in the useful signal area and the noise area is calculated as the intensity of the useful signal and the intensity of the noise, and the signal-to-noise ratio is calculated using the following formula. The formula is:

[0062] Signal-to-noise ratio = 10 × log 10 (Useful signal strength / noise strength)

[0063] Contrast refers to the degree of difference between brightness values ​​in an image, reflecting the difference in brightness between different areas in the image. To extract contrast, the original image can be converted into a grayscale image, traversing each pixel in the grayscale image, recording the maximum and minimum grayscale values ​​of the grayscale image, and calculating the difference between the maximum and minimum grayscale values. The obtained difference is the contrast.

[0064] Saturation refers to the purity or vividness of the color in the image. Saturation reflects the saturation of the color in the image, that is, the relative ratio of color to grayscale. To extract saturation, you can convert the original image into a grayscale image, traverse each pixel in the grayscale image, add the value of the saturation channel of each pixel to get a cumulative value, and divide the cumulative value by the total number of pixels to get the saturation of the image.

[0065] S203, performing dark light enhancement processing when at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation meets the first enhancement processing condition; after the dark light enhancement processing is completed, executing S204 or S205.

[0066] The method for satisfying the first enhancement processing condition is to set a value range or threshold for the brightness mean, the proportion of low brightness histogram, the signal-to-noise ratio, the contrast and the saturation. If the value of at least one of the brightness mean, the proportion of low brightness histogram, the signal-to-noise ratio, the contrast and the saturation is not within the value range or exceeds the threshold, it is determined that the first enhancement processing condition is satisfied; if the value of at least one of the brightness mean, the proportion of low brightness histogram, the signal-to-noise ratio, the contrast and the saturation is within the value range or does not exceed the threshold, it is determined that the first enhancement processing condition is not satisfied.

[0067] In the technical solution, optionally, the identification process that meets the first enhanced processing condition includes:

[0068] Constructing a feature vector composed of at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast, and saturation;

[0069] The feature vector is input into a pre-obtained machine learning model. If the machine learning model outputs a first result, it is determined that the first enhancement processing condition is met; if the machine learning model outputs a second result, it is determined that the first enhancement processing condition is not met.

[0070] In machine learning and data analysis, a feature vector refers to a vector that represents a data sample in a feature space. Each data sample can be represented as a feature vector. Each dimension of the feature vector corresponds to a feature or attribute of the data sample. Specifically, an array of {σ0, σ1, σ2, σ3, σ4} can be constructed as the feature vector of the original image, where σ0 represents the mean brightness, σ1 represents the proportion of low brightness histograms, σ2 represents the signal-to-noise ratio, σ3 represents the contrast, and σ4 represents the saturation. If the value of σ0, σ1, σ2, σ3, or σ4 is -1, it means that there is no data for this item. For example, if the feature vector of an original image is {208, -1, 153, 97, -1}, it means that the mean brightness of the original image is 208, there is no low brightness histogram proportion data, the signal-to-noise ratio is 153, the contrast is 97, and there is no saturation data.

[0071] The machine learning model is a mathematical model that learns from data and makes predictions or decisions through a machine learning algorithm. The machine learning model of this technical solution specifically adopts a decision tree model. The decision tree model is a machine learning model based on a tree structure, which is used for decision-making and classification tasks. The decision tree model performs a series of judgments or splits on the input features to ultimately achieve the purpose of classifying or predicting the sample.

[0072] The construction process of the decision tree model can be divided into two key steps: feature selection and tree construction. Feature selection determines which feature is selected at each node for judgment or splitting. Common feature selection methods include information gain, information gain ratio, and Gini index. These methods evaluate the classification ability or purity of samples based on features and select the best feature as the basis for node judgment. Tree construction is the process of generating a decision tree by recursively partitioning data. The specific steps include: selecting a feature as the current node and dividing the data set into different subsets based on the feature; for each subset, if the samples in the subset belong to the same category, the current node is marked as a leaf node, and the category is used as the output result of the leaf node; if the samples in the subset belong to different categories, the next feature is selected and partitioned to construct a child node; the above steps are repeated for the child node recursively until the stop condition is met, such as the number of samples in the node is less than the threshold or the depth of the tree reaches a predetermined value.

[0073] Specifically, the machine learning model is set to Γ, and the indicator calculation model is constructed:

[0074] Q=Γ(Ω)

[0075] Where, Ω={σ i}, i∈0,1,2,3,4, represents the feature vector; Q is the dark light enhancement index, Q=0 is the output of the first result, indicating that no enhancement is required, and Q=1 is the output of the second result, indicating that enhancement is required.

[0076] The advantage of this arrangement of the present scheme is that by inputting a feature vector including parameter information into a machine learning model and identifying whether the first enhancement processing condition is met based on the output result of the machine learning model, the accuracy and efficiency of the identification result of whether the first enhancement processing condition is met can be improved.

[0077] The dark light enhancement processing can be performed by increasing the brightness of the original image.

[0078] In this technical solution, optionally, dark light enhancement processing is performed, including:

[0079] Get the dark light enhancement coefficient set;

[0080] Polling the dark light enhancement coefficient set to obtain an enhanced image corresponding to each dark light enhancement coefficient;

[0081] According to the information entropy of each enhanced image, the target dark light enhancement coefficient is determined;

[0082] The target dark light enhancement coefficient is used to perform dark light enhancement processing to obtain an output image.

[0083] In the dark light enhancement process, the dark light enhancement coefficient is a parameter used to control the degree of image enhancement. Specifically, a larger dark light enhancement coefficient can achieve a more obvious enhancement effect and improve the visibility of dark details, but may introduce noise or over-enhancement problems; a smaller dark light enhancement coefficient will produce a slighter enhancement effect, maintain a low noise level, but may not completely improve dark details. Therefore, the maximum dark light enhancement coefficient K can be set, and the dark light enhancement coefficient set is (1, 2, 3...K-1, K), and the target dark light enhancement coefficient is determined in the dark light enhancement coefficient set according to the information entropy.

[0084] Polling is a computer programming technique used to periodically query or check whether a state or condition is met. Polling the dark light enhancement coefficient set, that is, keeping the original image I unchanged, changing the polling enhancement coefficient k, inputting the original image I and the polling enhancement coefficient k (k = 1 ~ K) into the following formula, and obtaining K enhanced images I', the formula is:

[0085]

[0086] Among them, A=-0.3293, B=1.1258.

[0087] Information entropy is a concept used in information theory to measure the uncertainty or amount of information of a random variable. In information theory, the information entropy of a random variable represents the average amount of information contained in the random variable. If a random variable has more possible values ​​and the probability of each value occurring is more uniform, the information entropy of the random variable will be greater. The unit of information entropy is bits. The calculation formula of information entropy is as follows:

[0088] Q=∑p i log2p i

[0089] Among them, Q represents information entropy, p i The pixel proportion or probability distribution of each gray level of the enhanced image I' is represented. The enhanced image I' is converted into a grayscale image, and the histogram statistics of the grayscale image are performed to calculate the number of pixels at each gray level. The number of pixels at each gray level is divided by the total number of pixels in the image to determine the pixel proportion or probability distribution p of each gray level. i .

[0090] The target dark light enhancement coefficient may be determined by selecting an enhancement coefficient k' with the largest information entropy Q as the target dark light enhancement coefficient.

[0091] The original image I and the target dark light enhancement coefficient k' are input into The enhanced image I' obtained at this time is the final output image.

[0092] The advantage of such a setting of the present scheme is that the output effect of the dark light enhancement processing can be improved by determining the target dark light enhancement coefficient according to the information entropy.

[0093] S204: When it is identified that the parameter information satisfies a second enhancement processing condition, perform a second enhancement processing.

[0094] S205: Perform contrast enhancement processing to obtain an enhancement processing result of the original image.

[0095] The advantage of this arrangement of the scheme is that by performing dark light enhancement processing on the original image, the dark details of the original image can be enhanced and the image quality can be improved.

[0096] Embodiment 3

[0097] Figure 3 It is a flow chart of the enhancement method of the image collected by unmanned equipment provided in the third embodiment of the present application. This scheme makes a better improvement on the above embodiment, and the specific improvement is: extracting the parameter information of the original image, including: dividing the original image according to a preset rule to obtain at least two image blocks; calculating the contrast of each image block; when it is identified that the parameter information meets the second enhancement processing condition, performing the second enhancement processing, including: when the variance of the contrast of each image block is less than a set threshold, determining that it meets the preset second enhancement processing condition, and performing defogging enhancement processing.

[0098] like Figure 3 As shown, the specific steps include:

[0099] S301, obtaining original images collected by unmanned equipment.

[0100] S302: Divide the original image according to a preset rule to obtain at least two image blocks.

[0101] An image block can be a rectangular area or a rectangular sub-image in the original image.

[0102] The preset rule can be to divide the original image into M*N rectangular blocks. For example, the original image is set to be divided into 5*4 rectangular blocks. If the length of the original image is 250 pixels and the width is 160 pixels, then the length of the rectangular block is 250 / 5=50 pixels and the width is 160 / 4=40 pixels.

[0103] By segmenting the original image according to preset rules, the original image can be divided.

[0104] S303: Calculate the contrast of each image block.

[0105] The Laplace enhancement method can be used to calculate the contrast. The Laplace enhancement method is an enhancement method based on the second-order derivative of the image. It improves the sharpness and contrast of the image and enhances the visibility of details by calculating the Laplace operator (second-order differential) of the image. The second-order differential is the process of taking the derivative of a function twice.

[0106] S304, when the variance of the contrast of each image block is less than the set threshold, it is determined that the preset second enhancement processing condition is met, and a defogging enhancement processing is performed; after the defogging enhancement processing is completed, S305 or S306 is executed.

[0107] Variance is a statistic used to measure the degree of dispersion or variation of a set of data or random variables. Variance measures the degree of deviation between each data point in the data set and the mean of the data set. The larger the variance, the greater the dispersion of the data points relative to the mean.

[0108] The steps of calculating the variance of the contrast of each image block include: adding the contrast of all image blocks and dividing by the number of image blocks to obtain the mean of the contrast; calculating the difference between the contrast of each image block and the mean, and squaring the difference; adding all square differences and dividing by the number of image blocks to obtain the variance of the contrast of each image block.

[0109] Setting the threshold requires comprehensive consideration of the image noise level, expected processing effect, and actual application scenario.

[0110] Dehazing enhancement is an image enhancement technique used to improve the quality of blurred and low-contrast images caused by atmospheric fog and / or haze, and can restore lost details and contrast in the image.

[0111] In this technical solution, optionally, defogging enhancement processing is performed, including:

[0112] The dark channel prior algorithm is used to perform dehazing and enhancement processing on the input image to obtain the output image.

[0113] The dark channel prior algorithm is a commonly used image dehazing algorithm. Due to the presence of atmospheric fog and / or haze, at least some pixels in the collected image have very low values ​​on at least one color channel, close to zero. Based on this principle, the dark channel prior algorithm uses the dark channel information of the image to estimate the atmospheric light and transmittance, thereby performing image dehazing.

[0114] First, calculate the dark channel image of the original image. The dark channel image is an image that selects the minimum value from the RGB channels of the image at each pixel position. Specifically, for each pixel position (x, y), the calculation is as follows:

[0115] Dark(x,y)=min(R(x,y),G(x,y),B(x,y))

[0116] Among them, Dark represents the dark channel image, R represents the red channel, G represents the green channel, and B represents the blue channel.

[0117] Then, the atmospheric light A is estimated by selecting the area with larger pixel value in the dark channel image.

[0118] Then, the transmittance t at each pixel position is calculated by using the estimated atmospheric light A. The transmittance represents the attenuation of light when passing through atmospheric haze, and the calculation formula is:

[0119] t(x,y)=1-w×min(Dark(x,y) / A)

[0120] Among them, w is a constant (usually 0.95) used to control the influence of transmittance.

[0121] Finally, the fogged image is repaired using the estimated transmittance t and atmospheric light A. For each pixel position (x, y) and each channel c (R, G, B), the repaired pixel value R'(x, y, c) is calculated using the following formula:

[0122] R'(x,y,c)

[0123] =(R(x, y, c)-A(c)) / max(t(x, y), t0)+A(c) where R(x, y, c) is the pixel value of the fogged image and t0 is a small positive number used to avoid division by zero.

[0124] The advantage of this setting of the scheme is that the calculation process of the dark channel prior algorithm is intuitive and simple, and the parameters can be adjusted according to specific needs to obtain better defogging effect.

[0125] S305: When it is identified that the parameter information satisfies a first enhancement processing condition, perform a first enhancement processing.

[0126] S306, performing contrast enhancement processing to obtain an enhancement processing result of the original image.

[0127] The advantage of this setting of the scheme is that by performing defogging and enhancement processing on the original image, the problem of low visibility of images collected by unmanned equipment during operations in foggy weather can be solved.

[0128] Embodiment 4

[0129] Figure 4This is a schematic diagram of the structure of an unmanned equipment image enhancement device provided in the fourth embodiment of the present invention. The device is used to execute the unmanned equipment image enhancement method described above, and has the corresponding functional modules and beneficial effects of the execution method. Figure 4 As shown, the device specifically includes:

[0130] The parameter information extraction module 410 is used to obtain the original image collected by the unmanned equipment and extract the parameter information of the original image;

[0131] The first enhancement processing module 420 is configured to perform the first enhancement processing when it is identified that the parameter information meets the first enhancement processing condition, and / or

[0132] The second enhancement processing module 430 is used to perform the second enhancement processing and the contrast enhancement processing to obtain the enhancement processing result of the original image when it is identified that the parameter information meets the second enhancement processing condition.

[0133] In the embodiment of the device, the parameter information extraction module is used to obtain the original image collected by the unmanned equipment and extract the parameter information of the original image; the enhancement processing module is used to perform the first enhancement processing when it is identified that the parameter information meets the first enhancement processing condition, and / or, when it is identified that the parameter information meets the second enhancement processing condition, perform the second enhancement processing, and perform contrast enhancement processing to obtain the enhancement processing result of the original image. This solution can solve the problem that the images collected by the unmanned equipment during agricultural operations are affected by factors such as weather and environment, resulting in poor image acquisition results and failure to normally reflect the growth status of crops, by performing the first enhancement processing and / or the second enhancement processing while performing the contrast enhancement processing, thereby achieving multi-faceted improvements in image quality.

[0134] Embodiment 5

[0135] Figure 5 A schematic diagram of the structure of an enhancement device for collecting images by unmanned equipment provided in an embodiment of the present invention, such as Figure 5 As shown, the device includes a processor 501, a memory 502, an input device 503 and an output device 504; the number of processors 501 in the device can be one or more. Figure 5 A processor 501 is taken as an example; the processor 501, the memory 502, the input device 503 and the output device 504 in the device can be connected by a bus or other means. Figure 5The example of connecting via a bus is taken. The memory 502, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as the program instructions / modules corresponding to the method for enhancing the image collected by unmanned equipment in the embodiment of the present invention. The processor 501 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 502, that is, realizes the above-mentioned mobile device target determination method. The input device 503 can be used to receive input digital or character information, and generate key signal input related to user settings and function control of the device. The output device 504 may include a display device such as a display screen.

[0136] Embodiment 6

[0137] Embodiment 6 of the present invention provides a storage medium including computer executable instructions, which may be stored in the form of a server application. When the computer executable instructions are executed by a computer processor, they are used to execute a method for enhancing images collected by unmanned equipment, the method comprising:

[0138] Acquire the original image collected by the unmanned equipment, and extract parameter information of the original image;

[0139] When it is identified that the parameter information meets the first enhancement processing condition, the first enhancement processing is performed, and / or, when it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and contrast enhancement processing is performed to obtain the enhancement processing result of the original image.

[0140] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0141] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal (which can be a mobile device, mobile phone, computer, server or network device, etc.) to execute the methods described in each embodiment of the present application.

[0142] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. An enhancement method for images collected by unmanned equipment, characterized in that: The method comprises: Acquire the original image collected by the unmanned equipment, and extract parameter information of the original image; When it is identified that the parameter information satisfies the first enhancement processing condition, the first enhancement processing is performed, and / or, When it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and the contrast enhancement processing is performed to obtain the enhancement processing result of the original image.

2. The method for enhancing images collected by unmanned equipment according to claim 1, characterized in that: Extracting parameter information of the original image includes: Extracting at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast and saturation of the original image; When it is identified that the parameter information satisfies the first enhancement processing condition, performing the first enhancement processing includes: When at least one of the image brightness mean, low brightness histogram proportion, signal-to-noise ratio, contrast and saturation meets the first enhancement processing condition, dark light enhancement processing is performed.

3. The method for enhancing images collected by unmanned equipment according to claim 2, characterized in that: The identification process that meets the first enhanced processing condition includes: Constructing a feature vector composed of at least one of the image brightness mean, low brightness histogram ratio, signal-to-noise ratio, contrast, and saturation; The feature vector is input into a pre-obtained machine learning model. If the machine learning model outputs a first result, it is determined that the first enhancement processing condition is met; if the machine learning model outputs a second result, it is determined that the first enhancement processing condition is not met.

4. The method for enhancing images collected by unmanned equipment according to claim 2, characterized in that: Perform dark light enhancement processing, including: Get the dark light enhancement coefficient set; Polling the dark light enhancement coefficient set to obtain an enhanced image corresponding to each dark light enhancement coefficient; According to the information entropy of each enhanced image, the target dark light enhancement coefficient is determined; The target dark light enhancement coefficient is used to perform dark light enhancement processing to obtain an output image.

5. The method for enhancing images collected by unmanned equipment according to claim 1, characterized in that: Extracting parameter information of the original image includes: Dividing the original image according to a preset rule to obtain at least two image blocks; Calculate the contrast of each image block; When it is identified that the parameter information satisfies the second enhancement processing condition, performing the second enhancement processing includes: When the variance of the contrast of each image block is less than a set threshold, it is determined that the preset second enhancement processing condition is met, and the defogging enhancement processing is performed.

6. The method for enhancing images collected by unmanned equipment according to claim 5, characterized in that: Perform enhanced defogging processing, including: The dark channel prior algorithm is used to perform dehazing and enhancement processing on the input image to obtain the output image.

7. The method for enhancing images collected by unmanned equipment according to claim 1, characterized in that: When it is identified that the parameter information meets the first enhancement processing condition, the first enhancement processing is performed, and / or, when it is identified that the parameter information meets the second enhancement processing condition, the second enhancement processing is performed, and the contrast enhancement processing is performed to obtain the enhancement processing result of the original image, including: When it is identified that the parameter information satisfies a first enhancement processing condition, performing a first enhancement processing; When the first enhancement processing result satisfies the second enhancement processing condition, performing the second enhancement processing to obtain the second enhancement processing result; Performing contrast enhancement processing on the second enhancement processing result to obtain an enhancement processing result of the original image; or, When it is identified that the parameter information satisfies the second enhancement processing condition, performing the second enhancement processing; When the second enhancement processing result satisfies the first enhancement processing condition, performing the first enhancement processing to obtain the first enhancement processing result; The first enhancement processing result is subjected to contrast enhancement processing to obtain an enhancement processing result of the original image.

8. An enhancement device for collecting images by unmanned equipment, characterized in that: The device comprises: A parameter information extraction module is used to obtain the original image collected by the unmanned equipment and extract the parameter information of the original image; a first enhancement processing module, configured to perform first enhancement processing when it is identified that the parameter information meets a first enhancement processing condition, and / or The second enhancement processing module is used to perform the second enhancement processing and the contrast enhancement processing to obtain the enhancement processing result of the original image when it is identified that the parameter information meets the second enhancement processing condition.

9. Unmanned equipment, including: one or more processors; A storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method for enhancing images collected by unmanned equipment as described in any one of claims 1-7.

10. A storage medium storing computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute the method for enhancing images collected by unmanned equipment according to any one of claims 1 to 7.

Images