Integrated image customization material sales intelligent terminal control system

Abstract

The invention relates to the technical field of integrated image customization and intelligent sales decision making, in particular to an integrated image customization material sales intelligent terminal control system, which comprises an intelligent terminal main control center used for calling to-be-printed image data and material inventory unit information selected by a user; the image reproduction complexity modeling unit is used for generating an image reproduction complexity index; the material medium interference potential analysis unit is used for generating a material medium interference potential; the process feasibility coupling calculation unit is used for calculating a process feasibility probability; the dynamic sales decision reconstruction unit is used for comparing the process feasibility probability with a preset reference qualified rate threshold value; if the process feasibility probability is smaller than a preset standard qualified rate threshold value, generating a dynamic risk premium instruction; otherwise, generating a basic selling price maintaining instruction; according to the method, the control strategy can be dynamically optimized along with equipment state aging or model change, and the problem that a static model is difficult to adapt to equipment dynamic performance change is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of integrated image customization and intelligent sales decision-making, in particular to an integrated image customization material sales intelligent terminal control system. BACKGROUND

[0002] The integrated image customization material sales intelligent terminal control system aims to solve the problem of how to realize dynamic and accurate sales pricing decision-making in image customization services according to the complexity of the image to be printed and the characteristics of the selected material, combined with the real-time state of the terminal device. In the prior art, the sales pricing of image customization materials may mainly rely on fixed cost accounting or experience rules, lacking quantitative evaluation of the printing process feasibility; this method is difficult to adapt to the varying complexity of the printing task, different image data have different precision requirements for the printing device, and high complexity images are prone to defects; the physical properties of the material hinder the image quality and affect the finished product pass rate; and the real-time state of the device determines the current process tolerance, and when the performance decreases, the risk of printing failure increases; due to the lack of a closed-loop control architecture of software and hardware collaboration, the existing system is difficult to calculate the process feasibility probability by coupling the characteristics of image, material and device; therefore, it is difficult to pre-evaluate the risk before printing and generate dynamic risk premium instructions in time to reasonably hedge the reprint cost caused by high difficulty tasks or poor device state, so as to ensure economic benefits and customer experience. SUMMARY

[0003] To solve the above technical problems, the present application provides an integrated image customization material sales intelligent terminal control system, specifically, the technical scheme of the present application comprises: An intelligent terminal master center for responding to user interaction instructions, retrieving image data to be printed and user-selected material inventory unit information; An image reproduction complexity modeling unit for feature extraction and entropy calculation of the image data to be printed, generating an image reproduction complexity index; A material medium interference potential analysis unit for physical property mapping of the material inventory unit information, generating a material medium interference potential; A process feasibility coupling calculation unit for obtaining a process tolerance index of the current terminal physical device, combining the image reproduction complexity index and the material medium interference potential, and using a logistic function to solve the process feasibility probability; A dynamic sales decision reconstruction unit for obtaining a preset reference pass rate threshold, comparing the process feasibility probability with the preset reference pass rate threshold; If the process feasibility probability is less than the preset reference pass rate threshold, a dynamic risk premium instruction is generated; If the process feasibility probability is greater than or equal to the preset reference qualified rate threshold, a basic selling price maintenance instruction is generated.

[0004] Optionally, the image reproduction complexity modeling unit performs the following steps: The image data to be printed is converted into a single-channel grayscale matrix, and the pixel intensity values of the single-channel grayscale matrix are normalized to a preset interval to generate a normalized image matrix; The normalized image matrix is calculated by a sliding window to obtain a local standard deviation matrix, and the arithmetic mean of the local standard deviation matrix is calculated to generate an image texture intensity global mean value; The arithmetic mean of the normalized image matrix is calculated to generate an image global average brightness value; Based on the image texture intensity global mean value, the image global average brightness value, and a preset minimum normal number, the signal-to-noise ratio reciprocal term is calculated; The Shannon entropy value of the normalized image matrix is calculated; The signal-to-noise ratio reciprocal term and the Shannon entropy value are weighted and summed to generate an image reproduction complexity index.

[0005] Optionally, the material medium interference potential analysis unit performs the following steps: The surface arithmetic mean roughness value and the substrate absorbance value corresponding to the material inventory unit information are obtained; The ratio of the surface arithmetic mean roughness value to a preset limit roughness constant is calculated to generate a roughness interference term; The ratio of the substrate absorbance value to a preset optical density limit constant is calculated to generate an optical density interference term; The roughness interference term and the optical density interference term are weighted and summed using a preset influence factor weight to generate a material medium interference potential.

[0006] Optionally, the process feasibility coupling calculation unit performs the following steps: Real-time physical sensor data of the current terminal physical device is collected, and a process tolerance index is generated based on the real-time physical sensor data mapping; The product of the image reproduction complexity index and the material medium interference potential is calculated to generate a coupling conflict value; The difference between the coupling conflict value and the process tolerance index is calculated to generate a net risk value; Based on the preset discriminant sensitivity coefficient and the net risk value, the process feasibility probability is calculated using a logistic function.

[0007] Optionally, the dynamic sales decision reconstruction unit performs the following steps: When the dynamic risk premium instruction is generated, the difference between the preset reference qualified rate threshold and the process feasibility probability is calculated; The product of the difference and a preset risk premium coefficient is calculated to generate a premium range; obtain a basic selling price corresponding to the material inventory unit information, and calculate a dynamic selling price based on the basic selling price and the premium range; When the basic selling price maintenance instruction is generated, the basic selling price is directly marked as the dynamic selling price.

[0008] Optionally, the real-time physical sensor data comprises nozzle clogging rate data or laser power data; When the nozzle clogging rate data is collected, a difference between the numerical value 1 and the nozzle clogging rate data is calculated to generate a process tolerance index; When the laser power data is collected, the laser power data is converted by using a preset mapping function to generate the process tolerance index.

[0009] Optionally, the method further comprises a weight self-adaptive adjustment module, and the weight self-adaptive adjustment module performs the following steps: identifying a printing device type of the current terminal physical device; If the printing device type is a thermal sublimation printer, the weight coefficient corresponding to the reciprocal of the signal-to-noise ratio is increased; If the printing device type is a laser engraving machine, the weight coefficient corresponding to the Shannon entropy value is increased.

[0010] Optionally, the preset influence factor weight setting step is as follows: identifying a process contact type of the current terminal physical device; If the process contact type is a contact process, the weight coefficient of the roughness interference term is set to be greater than the weight coefficient of the optical density interference term; If the process contact type is a non-contact inkjet process, the weight coefficient of the optical density interference term is set to be greater than the weight coefficient of the roughness interference term.

[0011] Compared with the prior art, the present application has the following beneficial effects: 1. The present system realizes quantitative evaluation and closed-loop control of printing risk; by constructing a mathematical model of image reproduction complexity and material medium interference potential, and combining with real-time process tolerance of the equipment, a process feasibility probability is calculated by using a logistic function coupling; this overcomes the blindness of the prior art which only relies on experience or fixed cost accounting, establishes a closed-loop architecture of software and hardware cooperation, and can accurately quantify the success rate of image-material-equipment matching before production, thereby effectively reducing the waste risk caused by blind production; 2. The system establishes a dynamic pricing mechanism based on objective process risk; not only assesses risk, but also builds a dynamic sales decision-making logic; by comparing the process feasibility probability with the benchmark pass rate threshold, automatically generating risk premium instructions for high-risk tasks, and maintaining the basic selling price for low-risk tasks; this mechanism can reasonably hedge potential reprinting costs caused by high-difficulty images or special materials, solve the problem of profit reversal in the traditional fixed pricing model in complex customization scenarios, and ensure business efficiency; 3. The system improves the resolution accuracy of heterogeneous materials and complex images; a multi-dimensional feature extraction method is used to evaluate the reproduction difficulty by integrating texture intensity and information entropy on the image side, and to map surface roughness and absorbance to physical interference potential on the material side; this fine-grained physical property mapping mechanism fully considers the hindering effect of material surface characteristics on image quality, avoids the one-sidedness of single-dimensional evaluation, and significantly improves the system's adaptability to various non-standard customized materials.

[0012] 4. The system has the ability of device state perception and model adaptive adjustment; by collecting real-time sensor data such as nozzle clogging rate or laser power, the process tolerance index is dynamically updated to reflect the real performance of the device; at the same time, the system can adaptively adjust the weight coefficients in the calculation model according to the device type and contact method; this ensures that the control strategy can be dynamically optimized with the aging of the device state or the change of the model, solving the problem that the static model is difficult to adapt to the dynamic performance change of the device. BRIEF DESCRIPTION OF DRAWINGS

[0013] The application will be further explained in conjunction with the accompanying drawings and embodiments: Figure 1 is a structural diagram of the system of the application. DETAILED DESCRIPTION

[0014] In order to make the purpose, technical scheme and advantages of the application clearer, the application will be further described in detail in conjunction with specific embodiments.

[0015] Example 1: Please refer to Figure 1 An integrated image customization material sales intelligent terminal control system, comprising: An intelligent terminal master center for responding to user interaction instructions, calling image data to be printed and user-selected material inventory unit information; An image reproduction complexity modeling unit for feature extraction and entropy calculation on the image data to be printed, generating an image reproduction complexity index; A material medium interference potential analysis unit for physical property mapping on the material inventory unit information, generating a material medium interference potential; The process feasibility coupling calculation unit is used to obtain the process tolerance index of the current terminal physical equipment, and combine the image reproduction complexity index and the material medium interference potential to calculate the process feasibility probability using the logistic function. The dynamic sales decision reconstruction unit is used to obtain a preset benchmark pass rate threshold and compare the probability of process feasibility with the preset benchmark pass rate threshold. If the probability of process feasibility is less than the preset benchmark pass rate threshold, a dynamic risk premium instruction is generated. If the probability of process feasibility is greater than or equal to the preset benchmark pass rate threshold, a basic selling price maintenance instruction is generated.

[0016] An integrated intelligent terminal control system for customized image-based material sales is implemented through a closed-loop control architecture that integrates hardware and software. The system includes an intelligent terminal main control center, an image reproduction complexity modeling unit, a material medium interference potential analysis unit, a process feasibility coupling calculation unit, and a dynamic sales decision reconstruction unit. The intelligent terminal main control center is equipped with a data bus interface to respond to user interaction commands and retrieve the image data to be printed and the user-selected material inventory unit information in real time. The image reproduction complexity modeling unit performs feature extraction and entropy calculation on the image data to be printed, generating a dimensionless image reproduction complexity index. This index characterizes the level of image accuracy required by printing equipment; the material medium interference potential analysis unit maps the physical properties of the material inventory unit information to generate the material medium interference potential. This characterizes the hindering effect of material surface properties on imaging quality; the process feasibility coupling calculation unit obtains the process tolerance index of the current terminal physical equipment. Combined with the image reproduction complexity index Interference potential with material medium The probability of process feasibility is calculated using the logistic function. The dynamic sales decision reconstruction unit obtains the preset benchmark pass rate threshold. The probability of process feasibility Compare with this threshold; if the probability of process feasibility is... Less than the preset benchmark pass rate threshold Generate dynamic risk premium orders; if the probability of process feasibility... Greater than or equal to the preset benchmark pass rate threshold Generate a base price maintenance order.

[0017] Example 2: The image reproduction complexity modeling unit performs the following steps: The image data to be printed is converted into a single-channel grayscale matrix, and the pixel intensity values ​​of the single-channel grayscale matrix are normalized to a preset range to generate a normalized image matrix. The local standard deviation matrix is calculated by sliding window calculation on the normalized image matrix, the arithmetic mean of the local standard deviation matrix is calculated, and the global mean of the image texture intensity is generated; The arithmetic mean of the normalized image matrix is counted, and the global average brightness value of the image is generated; Based on the global mean of the image texture intensity, the global average brightness value of the image, and the preset minimum normal number, the signal-to-noise ratio reciprocal term is calculated; The Shannon entropy value of the normalized image matrix is calculated; The signal-to-noise ratio reciprocal term and the Shannon entropy value are weighted and summed to generate the image reproduction complexity index.

[0018] The image reproduction complexity modeling unit converts the image data to be printed into a single-channel gray matrix , and the pixel intensity value of the matrix is normalized to a preset interval to generate a normalized image matrix; the local standard deviation matrix is calculated by sliding window calculation on the normalized image matrix , the arithmetic mean of the local standard deviation matrix is calculated, and the global mean of the image texture intensity is generated ; the arithmetic mean of the normalized image matrix is counted, and the global average brightness value of the image is generated ; based on the global mean of the image texture intensity , the global average brightness value of the image , and the preset minimum normal number , the signal-to-noise ratio reciprocal term is calculated; the Shannon entropy value of the normalized image matrix is calculated ; the signal-to-noise ratio reciprocal term and the Shannon entropy value are weighted and summed to generate the image reproduction complexity index ; The values of the weight coefficients and are determined by a pre-constructed calibration process; a calibration data set containing standard sample images is constructed , the actual defect rate of each sample image under standard printing conditions is recorded , and the corresponding calibration texture intensity and calibration entropy value ; the mapping relationship between , and is established by multiple regression analysis, so as to calculate the values of and that minimize the model prediction error, and the specific calculation formula is as follows:

[0019] , wherein takes the value of to prevent the denominator from being zero.

[0020] Embodiment 3: The material medium interference potential analysis unit performs the following steps: Obtain the surface arithmetic average roughness value and the substrate absorbance value corresponding to the material inventory unit information; Calculate the ratio of the surface arithmetic average roughness value to the preset limit roughness constant to generate a roughness interference term; Calculate the ratio of the substrate absorbance value to the preset optical density limit constant to generate an optical density interference term; Weighted sum the roughness interference term and the optical density interference term using the preset influence factor weight to generate the material medium interference potential.

[0021] The material medium interference potential analysis unit obtains the surface arithmetic average roughness value corresponding to the material inventory unit information from the preset database and the substrate absorbance value ; calculates the ratio of the surface arithmetic average roughness value to the preset limit roughness constant to generate a roughness interference term; calculates the ratio of the substrate absorbance value to the preset optical density limit constant to generate an optical density interference term; and weighted sums the roughness interference term and the optical density interference term using the preset influence factor weight and to generate the material medium interference potential , the specific calculation formula is as follows:

[0022] is the roughness physical limit allowed by the current printing technology, is the optical density limit constant.

[0023] Embodiment 4: The process feasibility coupling calculation unit performs the following steps: Collect real-time physical sensor data of the current terminal physical device, and generate a process tolerance index based on the real-time physical sensor data mapping; Calculate the product of the image reproduction complexity index and the material medium interference potential to generate a coupling conflict value; Calculate the difference between the coupling conflict value and the process tolerance index to generate a net risk value; Based on the preset discriminant sensitivity coefficient and the net risk value, calculate the process feasibility probability using the logistic function.

[0024] The process feasibility coupling calculation unit collects real-time physical sensor data of the current terminal physical device, and generates a process tolerance index based on the real-time physical sensor data mapping , which is strictly limited in the range of ; In order to ensure the unity of the calculation dimension, the system introduces a preset system load normalization coefficient ; the image reproduction complexity index is calculated as the product of the material medium interference potential ; and the system load normalization coefficient is used to scale the product to generate a normalized load index ; the normalized load index is used to represent the pressure level of the current printing task relative to the theoretical limit of the device; The difference between the normalized load index and the process tolerance index is calculated to generate a net risk value ; based on the preset discriminant sensitivity coefficient and the net risk value , the process feasibility probability is calculated using the logistic function ; this calculation process ensures the accuracy of the probability evaluation model in the nonlinear response region of the device performance by unifying the task load and device capability to the same dimension interval, and the specific calculation formula is as follows:

[0025]

[0026] In the formula, is the system load normalization coefficient, which is the inverse of the system theoretical maximum load value , wherein According to the image data bit depth and texture limit calculation, specifically, is the weighted theoretical maximum value, and the calculation formula is ; for an 8-bit grayscale image, the theoretical maximum entropy value is 8, and the maximum texture intensity is 0.5, is the sum of the material interference potential weight coefficients ; this coefficient is used to map the task load to space to match the dimension of the process tolerance index; is a dimensionless constant that controls the steepness of the S-shaped curve, which is determined by the inherent error conversion bandwidth of the device.

[0027] Embodiment 5: The dynamic sales decision reconstruction unit performs the following steps: When the dynamic risk premium instruction is generated, a difference between the preset benchmark qualified rate threshold and the process feasibility probability is calculated; A product of the difference and a preset risk premium coefficient is calculated to generate a premium range; The basic selling price corresponding to the material inventory unit information is obtained, and a dynamic selling price is calculated based on the basic selling price and the premium range; When the basic selling price maintenance instruction is generated, the basic selling price is directly marked as the dynamic selling price.

[0028] When the dynamic risk premium instruction is generated, the dynamic sales decision reconstruction unit calculates a difference between a preset benchmark qualified rate threshold and the process feasibility probability ; a product of the difference and a preset risk premium coefficient is calculated to generate a premium range; the basic selling price corresponding to the material inventory unit information is obtained , and a dynamic selling price is calculated based on the basic selling price and the premium range ; when the basic selling price maintenance instruction is generated, the basic selling price is directly marked as the dynamic selling price , in order to ensure the unity of the calculation logic and prevent price inversion of high-quality prints, a boundary limiting function is introduced in the specific calculation formula as follows:

[0029] The formula ensures that when the process feasibility probability is greater than or equal to the benchmark threshold , the premium item is automatically zeroed, and the system only outputs the basic selling price; wherein, The coefficient is set based on the ratio of the historical average reprint cost to the average single price.

[0030] Embodiment 6: The real-time physical sensor data includes nozzle clogging rate data or laser power data; When the nozzle clogging rate data is collected, a difference between the numerical value 1 and the nozzle clogging rate data is calculated to generate a process tolerance index; When the laser power data is collected, the laser power data is converted using a preset mapping function to generate a process tolerance index.

[0031] The real-time physical sensor data includes nozzle clogging rate data or laser power data ; when the nozzle clogging rate data is collected, a difference between the numerical value 1 and the nozzle clogging rate data is calculated to generate a process tolerance index , and a preset mapping function is configured as a negative correlation mapping relationship, that is, laser power data The higher the value is, the closer the device output power is to the limit or the laser is seriously aged, and the process tolerance index is lower; on the contrary, the lower the power value is, the device is in a light load and high efficiency interval, and the process tolerance index is higher; when the laser power data is collected, the preset mapping function is used to convert the laser power data to generate the process tolerance index .

[0032] Embodiment 7: The system also includes a weight adaptive adjustment module, which performs the following steps: Identify the print device type of the current terminal physical device; If the print device type is a thermal sublimation printer, increase the weight coefficient corresponding to the reciprocal of the signal-to-noise ratio term; If the print device type is a laser engraving machine, increase the weight coefficient corresponding to the Shannon entropy value.

[0033] The weight adaptive adjustment module identifies the print device type of the current terminal physical device; if the print device type is a thermal sublimation printer, the weight coefficient corresponding to the reciprocal of the signal-to-noise ratio term is increased to increase the suppression weight of low signal-to-noise ratio features; if the print device type is a laser engraving machine, the weight coefficient corresponding to the Shannon entropy value is increased to focus on information amount evaluation.

[0034] Embodiment 8: The setting steps of the preset influence factor weight are as follows: Identify the process contact type of the current terminal physical device; If the process contact type is a contact process, set the weight coefficient of the roughness interference term to be greater than the weight coefficient of the optical density interference term; If the process contact type is a non-contact inkjet process, set the weight coefficient of the optical density interference term to be greater than the weight coefficient of the roughness interference term.

[0035] The setting of the preset influence factor weight and is determined according to the process contact type of the current terminal physical device; if the process contact type is a contact process, the weight coefficient of the roughness interference term is set to be greater than the weight coefficient of the optical density interference term ; if the process contact type is a non-contact inkjet process, the weight coefficient of the optical density interference term is set to be greater than the weight coefficient of the roughness interference term . ​.

[0036] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the present application.

Claims

1. An integrated image customization material sales intelligent terminal control system, characterized in that, The method comprises the following steps: an intelligent terminal master center is used to respond to user interaction instructions, call image data to be printed and user-selected material inventory unit information; an image reproduction complexity modeling unit is used to extract features and calculate entropy values of the image data to be printed, and generate an image reproduction complexity index; a material medium interference potential analysis unit is used to map physical properties of the material inventory unit information, and generate a material medium interference potential; a process feasibility coupling calculation unit is used to obtain a process tolerance index of a current terminal physical device, combine the image reproduction complexity index and the material medium interference potential, and use a logistic function to calculate a process feasibility probability; a dynamic sales decision reconstruction unit is used to obtain a preset benchmark pass rate threshold, compare the process feasibility probability with the preset benchmark pass rate threshold, and generate a dynamic risk premium instruction if the process feasibility probability is less than the preset benchmark pass rate threshold, or generate a basic sales price maintenance instruction if the process feasibility probability is greater than or equal to the preset benchmark pass rate threshold. The image reproduction complexity modeling unit performs the following steps: convert the image data to be printed into a single-channel grayscale matrix, normalize pixel intensity values of the single-channel grayscale matrix to a preset interval, and generate a normalized image matrix; 2. The integrated image customization material selling intelligent terminal control system according to claim 1, wherein, perform sliding window calculation on the normalized image matrix to obtain a local standard deviation matrix, calculate an arithmetic mean value of the local standard deviation matrix, and generate an image texture intensity global mean value; calculate an arithmetic mean value of the normalized image matrix, and generate an image global average brightness value; calculate a signal-to-noise ratio reciprocal term based on the image texture intensity global mean value, the image global average brightness value, and a preset minimum normal number; calculate a Shannon entropy value of the normalized image matrix; perform weighted summation on the signal-to-noise ratio reciprocal term and the Shannon entropy value, and generate the image reproduction complexity index. The material medium interference potential analysis unit performs the following steps: obtain a surface arithmetic mean roughness value and a substrate absorbance value corresponding to the material inventory unit information; 3. The integrated image customization merchandise sales intelligent terminal control system according to claim 1, characterized in that, calculate a ratio of the surface arithmetic mean roughness value to a preset limit roughness constant, and generate a roughness interference term; calculate a ratio of the substrate absorbance value to a preset optical density limit constant, and generate an optical density interference term; perform weighted summation on the roughness interference term and the optical density interference term by using a preset influence factor weight, and generate the material medium interference potential. The process feasibility coupling calculation unit performs the following steps: collect real-time physical sensor data of the current terminal physical device, and map the real-time physical sensor data to generate the process tolerance index; 4. The integrated image customization merchandise sales intelligent terminal control system according to claim 1, characterized in that, calculate a product of the image reproduction complexity index and the material medium interference potential, and generate a coupling conflict value; calculate a difference value between the coupling conflict value and the process tolerance index, and generate a net risk value; calculate the process feasibility probability by using a logistic function based on a preset discriminant sensitivity coefficient and the net risk value. The dynamic sales decision reconstruction unit performs the following steps: when the dynamic risk premium instruction is generated, calculate a difference value between the preset benchmark pass rate threshold and the process feasibility probability; 5. The integrated image customization merchandise sales intelligent terminal control system according to claim 1, characterized in that, calculate a product of the difference value and a preset risk premium coefficient, and generate a premium amplitude; obtain a basic sales price corresponding to the material inventory unit information, and calculate a dynamic sales price based on the basic sales price and the premium amplitude. ​ ​ When the base selling price maintenance instruction is generated, the base selling price is directly marked as the dynamic selling price.

6. The integrated image customization merchandise vending kiosk control system of claim 4, wherein, The real-time physical sensor data includes nozzle clogging rate data or laser power data; When the nozzle clogging rate data is collected, the difference between the numerical value 1 and the nozzle clogging rate data is calculated to generate a process tolerance index; When the laser power data is collected, the laser power data is converted using a preset mapping function to generate a process tolerance index.

7. The integrated image customization merchandise vending kiosk control system of claim 2, wherein, The weight adaptive adjustment module further includes the following steps: Identify the printing device type of the current terminal physical device; If the printing device type is a thermal sublimation printer, increase the weight coefficient corresponding to the reciprocal of the signal-to-noise ratio; If the printing device type is a laser engraving machine, increase the weight coefficient corresponding to the Shannon entropy value.

8. The integrated image customization merchandise vending kiosk control system of claim 3, wherein, The preset influence factor weight setting steps are as follows: Identify the process contact type of the current terminal physical device; If the process contact type is a contact process, set the weight coefficient of the roughness interference term to be greater than the weight coefficient of the optical density interference term; If the process contact type is a non-contact inkjet process, set the weight coefficient of the optical density interference term to be greater than the weight coefficient of the roughness interference term.

Images