Adaptive Control Method for Imaging Parameters of Photoelectric Sorting System Based on Real-Time Feedback

Abstract

This invention discloses an adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback, relating to the field of photoelectric sorting technology. The method includes: acquiring imaging region feedback data and illumination unit light attenuation feedback data; separating locally bright areas and non-bright normal areas from the imaging region feedback data; and forming a bright spot feedback traction quantity based on the brightness change, area ratio, and overall brightness feedback change of the locally bright areas; and determining the correspondence between the brightness change direction of the locally bright areas and the overall brightness feedback change direction of the imaging region based on the bright spot feedback traction quantity to form a bright spot feedback traction result. This invention, by forming the bright spot feedback traction quantity based on the brightness change, area ratio, and overall brightness feedback change of the locally bright areas, can quantitatively determine the traction effect of locally bright areas on the overall brightness feedback change direction of the imaging region.

Description

Technical Field

[0001] This invention relates to the field of photoelectric sorting technology, specifically to an adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback. Background Technology

[0002] Photoelectric sorting systems typically acquire image data of the material to be sorted within the imaging area using an imaging device. Combined with stable illumination provided by an illumination unit, the system identifies the material's color, brightness, texture, outline, and surface condition, thus providing a basis for subsequent sorting. In actual operation, there is a correlation between the feedback data from the imaging area and the light output state of the illumination unit. Changes in illumination brightness directly affect the image brightness response, thereby impacting the stability of target material identification. Therefore, adjusting the illumination brightness based on real-time feedback data is a crucial way to improve the operational consistency of photoelectric sorting systems. In the existing photoelectric sorting control process, the system usually judges whether the current image is too bright or too dim based on the overall brightness feedback change of the imaging area, and further generates illumination brightness adjustment instructions; at the same time, some systems will also combine the light attenuation feedback data of the illumination unit itself to compensate for the output attenuation of the illumination unit in order to maintain the illumination stability in the imaging area; the above methods can provide an effective basis for adjustment under normal brightness fluctuation and illumination attenuation scenarios, so that the image acquisition state is kept within a range suitable for recognition. However, when there are localized areas of strong reflection on the surface of the material to be sorted, these localized bright areas may influence the overall brightness feedback change of the imaging area in a short period of time, causing the overall brightness feedback change of the imaging area to appear overly bright. Meanwhile, the light attenuation feedback data from non-bright normal areas and the illumination unit may still indicate a need to maintain or increase the illumination brightness. If the brightness reduction feedback corresponding to the localized bright areas is directly written into the global illumination brightness control set, the illumination brightness control command may deviate from the actual imaging requirements of non-bright normal areas. Therefore, this invention proposes an adaptive control method for imaging parameters of the photoelectric sorting system based on real-time feedback. Summary of the Invention

[0003] The purpose of this invention is to provide an adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback, so as to solve the problems mentioned in the background art.

[0004] This invention can be achieved through the following technical solution: an adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback, comprising: Step 1: Obtain imaging area feedback data and illumination unit light attenuation feedback data. Separate local bright areas and non-bright normal areas from the imaging area feedback data. Based on the brightness change, area ratio of the local bright areas and the overall brightness feedback change of the imaging area, form the bright spot feedback traction amount. Step 2: Based on the brightness feedback traction amount, determine the corresponding brightness change direction of the local bright area and the overall brightness feedback change direction of the imaging area to form the brightness feedback traction result; Step 3: Based on the bright spot feedback traction results, perform bright spot subtraction processing on the overall brightness feedback change of the imaging area to form bright spot-removed brightness feedback data; Step 4: Based on the brightness feedback data of the bright spots, the regional brightness change of the local bright areas, and the light attenuation feedback data of the lighting units, determine the brightness feedback direction corresponding to the non-bright normal areas, the brightness reduction feedback direction of the local bright areas, and the light attenuation compensation feedback direction of the lighting units, respectively, and compare the directional relationships to form the feedback direction splitting result. Step 5: Generate bright spot reduction feedback stripping results based on feedback direction splitting results; according to the bright spot reduction feedback stripping results, do not write the local high brightness area reduction feedback direction into the global lighting brightness control basis set, and write the bright spot reduction brightness feedback data and lighting unit light attenuation feedback data into the global lighting brightness control basis set to generate lighting brightness control instructions.

[0005] A further technical improvement of the present invention is that step three, generating brightness feedback data for removing bright spots, includes: When the direction of the overall brightness feedback change of the imaging area is determined to be dominated by the local bright area based on the bright spot feedback traction result, the brightness deduction amount corresponding to the local bright area is determined based on the regional brightness change amount and area ratio of the local bright area, and the brightness deduction amount is deducted from the overall brightness feedback change amount of the imaging area to form bright spot removal brightness feedback data. When the direction of the overall brightness feedback change of the imaging area is not determined based on the brightness spot feedback traction result, the amount of overall brightness feedback change of the imaging area is used as the brightness feedback data for removing the brightness spot.

[0006] A further technical improvement of the present invention is that: in step five, the bright spot reduction feedback stripping result is formed based on the feedback direction splitting result, including: Based on the positional migration relationship, regional overlap relationship, and number of continuous control cycles of the local bright areas, a local transient bright spot determination result is formed. Within a preset verification control period after a locally bright area exits the imaging area, it is determined whether the brightness feedback direction of the non-bright normal area continues the brightness reduction feedback direction of the locally bright area, thus forming a representative migration judgment result of the imaging area. When the local transient bright spot determination result determines that the local high-brightness area belongs to the local transient bright spot that moves with the material, the imaging area representative migration determination result determines that the brightness feedback direction corresponding to the non-high-brightness normal area does not continue the brightness reduction feedback direction of the local high-brightness area, and the feedback direction split result determines that the brightness reduction feedback direction of the local high-brightness area is split from the brightness feedback direction corresponding to the non-high-brightness normal area and the light attenuation compensation feedback direction of the illumination unit, a bright spot brightness reduction feedback stripping result is formed.

[0007] A further technical improvement of the present invention is that the step of separating the locally bright area and the non-bright normal area from the imaging area feedback data in step one includes: In the imaging area feedback data, the pixel position and brightness response value of the material to be sorted are extracted based on the difference in brightness response between the pixel position and the background area. The contour boundary position is determined based on the pixel position of the material to be sorted, and the contour boundary position is used to enclose the material contour coverage position. Within the area covered by the material outline, multiple pixel positions are selected as high-brightness response positions in descending order of the brightness response values ​​corresponding to each pixel position, and the high-brightness response positions that satisfy the pixel adjacency relationship are connected to form a local high-brightness candidate area. During multiple imaging sampling moments when the same material to be sorted passes through the imaging area, the area and boundary position of the local bright candidate region are obtained respectively. The position of the local bright candidate region in the material outline coverage position is converted into the relative region position. According to the order of multiple imaging sampling moments, the region area change sequence, the region boundary change sequence, and the relative region position sequence are formed. When the region area change sequence and the region boundary change sequence meet the preset reflection shape change conditions, and the relative region position sequence corresponds to the same local contour position range within the material contour coverage position, the local highlight candidate region is determined as the local highlight region. The set of pixel locations within the material outline coverage area that do not fall into the local bright area, do not reach the saturation response state, and whose brightness response value is within the preset effective response range are determined as the non-bright normal area.

[0008] A further technical improvement of the present invention is that: the formation of the bright spot feedback traction in step one includes: The non-highlight normal area is divided into multiple normal feedback sub-regions, and the brightness change direction of each sub-region is calculated to form a set of normal feedback directions. The direction of brightness change in the local bright area is determined based on the change in brightness of the local bright area. The consistency of the direction of brightness change in the local bright area with the direction of brightness change in each sub-region in the set of normal feedback directions is compared. The proportion of normal feedback sub-regions that are inconsistent with the direction of brightness change in the local bright area is counted to form the local direction deviation. Calculate the proportion of the brightness change of the local bright area in the overall brightness feedback change of the imaging area, and form an area inverse vector based on the inverse value of the area proportion of the local bright area. By mapping the local directional deviation, the proportion of brightness contribution, and the area inverse vector to the same value range, we obtain the local directional deviation mapping value, the brightness contribution mapping value, and the area inverse mapping value. The local directional deviation mapping value, brightness contribution mapping value, and area reverse mapping value are weighted and summed according to the preset traction weights to form the bright spot feedback traction amount.

[0009] A further technical improvement of the present invention is that the step of forming the bright spot feedback traction result in step two includes: The traction order of the local bright areas is determined according to the order of the bright spot feedback traction amount from high to low. When the direction of brightness change in a local bright area is consistent with the direction of brightness feedback change in the overall imaging area, and the traction sorting position is within the preset traction sorting range, the local bright area is determined to be in the dominant traction state, and the local bright area is determined to be the dominant traction area. The brightness deduction amount corresponding to the dominant traction area is determined based on the regional brightness change and area ratio of the dominant traction area. The region location of the dominant traction area, the amount of regional brightness change, the brightness deduction, the traction sorting position, and the dominant traction status are written into the bright spot feedback traction results.

[0010] A further technical improvement of the present invention is that: after generating the brightness feedback data for removing bright spots in step three, the brightness feedback data for removing bright spots is corrected, including: Read the region location and brightness deduction amount corresponding to the dominant traction region from the bright spot feedback traction results, and compare the overlapping regions corresponding to each dominant traction region to determine the overlapping and non-overlapping regions. The brightness deduction amounts corresponding to the overlapping area positions are merged according to the rule of retaining the same pixel position once to obtain the overlapping area deduction amount, and the brightness deduction amounts corresponding to the non-overlapping area positions are determined as the non-overlapping area deduction amount. The deduction amounts for overlapping areas and non-overlapping areas are summed to form the deduplication amount; The deduplication amount is subtracted from the overall brightness feedback change in the imaging area to form the deduplication residual amount; Calculate the brightness change of the non-highlight normal area, and form a residual verification range based on the brightness change of the non-highlight normal area and the preset brightness deviation range; The residual amount is compared with the residual verification range. When the residual amount is within the residual verification range, the residual amount is determined as the verified brightness feedback data after removing bright spots. When the deducted residual amount is not within the residual verification range, the deducted residual amount is limited and corrected according to the boundary value closest to the deducted residual amount in the residual verification range, thus forming the corrected brightness feedback data for removing bright spots.

[0011] A further technical improvement of the present invention is that: the step four in forming the feedback direction splitting result includes: The brightness feedback direction corresponding to the non-high-brightness normal area is compared with the light attenuation compensation feedback direction of the lighting unit to form the result corresponding to the normal compensation direction. When the result of the normal compensation direction determines that the brightness feedback direction corresponding to the non-high brightness normal area is consistent with the light attenuation compensation feedback direction of the lighting unit, the feedback direction corresponding to the consistency between the two is extracted, and the feedback direction is determined as the normal compensation reference direction. The brightness reduction feedback direction of the local bright area is compared with the normal compensation reference direction. When the brightness reduction feedback direction of the local bright area is inconsistent with the normal compensation reference direction, the local bright area is identified as the direction split source area, and a direction split mark corresponding to the direction split source area and the brightness reduction feedback direction of the local bright area is generated. Write the brightness feedback direction corresponding to the non-high-brightness normal area, the light attenuation compensation feedback direction of the illumination unit, the brightness reduction feedback direction of the local high-brightness area, the area location corresponding to the direction split source area, and the direction split mark into the feedback direction split result.

[0012] Compared with the prior art, the present invention has the following beneficial effects: This invention separates locally bright areas and non-bright normal areas from the feedback data of the imaging area, and forms a bright spot feedback traction amount based on the brightness change, area ratio of the locally bright areas and the overall brightness feedback change of the imaging area. This allows for a quantitative judgment of the traction effect of locally bright areas on the direction of the overall brightness feedback change of the imaging area. Therefore, before forming the basis for illumination brightness control, it is necessary to first identify whether the locally bright areas dominate the direction of the overall brightness feedback change, avoiding the direct generation of global brightness reduction control basis based solely on the overall brightness feedback change of the imaging area. Furthermore, this invention performs bright spot subtraction processing on the overall brightness feedback change of the imaging area based on the bright spot feedback traction result, forming bright spot-removed brightness feedback data. It also verifies or limits the subtraction result by combining the regional brightness change of non-high-brightness normal areas and the residual verification range, so that the data involved in subsequent illumination brightness control can be closer to the brightness feedback state of non-high-brightness normal areas. This can reduce the influence of local high-brightness areas on the global brightness feedback judgment and improve the data consistency in the imaging parameter control process. On the other hand, based on the brightness feedback data of bright spots, the regional brightness change of local bright areas, and the light attenuation feedback data of the illumination unit, the present invention further forms the brightness feedback direction corresponding to the non-bright normal area, the brightness reduction feedback direction of the local bright area, and the light attenuation compensation feedback direction of the illumination unit, respectively, and forms the feedback direction splitting result by comparing the direction relationship; when the feedback direction splitting result shows that the brightness reduction feedback direction of the local bright area is separated from the normal brightness feedback requirement and the light attenuation compensation requirement of the illumination unit, the bright spot brightness reduction feedback stripping result is formed, so that the brightness reduction feedback direction of the local bright area is not written into the global illumination brightness control basis set, thereby improving the matching between the illumination brightness control command and the actual imaging recognition requirements. Attached Figure Description

[0013] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0014] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0015] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0016] Example 1

[0017] Please see Figure 1 As shown, this invention provides an adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback, including: Step 1: Obtain imaging area feedback data and illumination unit light attenuation feedback data. Separate local bright areas and non-bright normal areas from the imaging area feedback data. Based on the brightness change, area ratio of the local bright areas and the overall brightness feedback change of the imaging area, form the bright spot feedback traction amount. Specifically, in step one, imaging area feedback data and illumination unit light attenuation feedback data are acquired first. Imaging area feedback data is collected by the imaging device as the material to be sorted passes through the imaging area. This data includes the position of each pixel within the imaging area, the corresponding brightness response value, the imaging sampling time, and the control cycle identifier. Illumination unit light attenuation feedback data is collected by the light output detection component of the illumination unit. This data includes the current light output intensity, the reference light output intensity, and the light attenuation amount. Imaging area feedback data and illumination unit light attenuation feedback data within the same control cycle are correlated using the control cycle identifier. Brightness response values, area brightness values, area brightness changes, and overall brightness feedback changes in the imaging area are all represented using the same brightness scale. This brightness scale is determined by the imaging device's lowest effective brightness response value and highest unsaturated brightness response value, ensuring that subsequent difference calculations and ratio calculations are based on the same data caliber.

[0018] After obtaining the imaging area feedback data, the background area is first determined based on the empty imaging data. Specifically, empty imaging data is collected when no material to be sorted passes through the imaging area, and the pixel positions and their corresponding background brightness response values ​​in the empty imaging data are used as background area data. Within the current control cycle, the difference between the brightness response values ​​corresponding to each pixel position in the imaging area feedback data and the background brightness response values ​​corresponding to the same pixel positions in the background area data is calculated to form the brightness response difference corresponding to each pixel position. The fluctuation of the background brightness response values ​​at the same pixel positions in multiple empty imaging data is statistically analyzed to obtain the upper limit of background brightness fluctuation; the brightness difference between the material pixel positions and the background pixel positions in the material imaging data under the baseline sorting state is statistically analyzed to obtain the lower limit of material imaging difference; a preset material response difference condition is determined based on the upper limit of background brightness fluctuation and the lower limit of material imaging difference. When the brightness response difference corresponding to a certain pixel position meets the preset material response difference condition, the pixel position is determined as the pixel position corresponding to the material to be sorted, and the brightness response value corresponding to that pixel position is recorded.

[0019] After determining the pixel positions corresponding to the materials to be sorted, the contour boundary positions are determined based on these pixel positions. Specifically, for any pixel position corresponding to a material to be sorted, it is determined whether the adjacent pixel positions belong to the background area; when at least one adjacent pixel position belongs to the background area, the pixel position is determined as the contour boundary position; the contour boundary positions are connected according to the pixel adjacency relationship between them to form an enclosing contour; the pixel positions inside the enclosing contour and the pixel positions on the enclosing contour are jointly determined as the material contour coverage position.

[0020] Within the area covered by the material outline, the pixel positions are sorted in descending order of their corresponding brightness response values. Pixel positions located within a preset high-brightness sorting range and with brightness response values ​​higher than a preset high-brightness response threshold are identified as high-brightness response positions. The preset high-brightness sorting range is determined by the distribution of pixel positions in the reflection region of the local reflection sample data, and the preset high-brightness response threshold is determined by the lowest quantile of the brightness response values ​​of the pixels in the reflection region of the local reflection sample data. Subsequently, it is determined whether the high-brightness response positions satisfy pixel adjacency relationships. In this embodiment, pixel adjacency relationship means that the pixel distance between two high-brightness response positions does not exceed a preset pixel adjacency distance. The preset pixel adjacency distance is determined by the physical size of a single pixel in the imaging device and the maximum allowable break distance in the local reflection region. High-brightness response positions that satisfy pixel adjacency relationships are connected and merged to form local high-brightness candidate regions. Each local high-brightness candidate region consists of a set of high-brightness response positions, and the corresponding region boundary pixel positions and region internal pixel positions are recorded.

[0021] At multiple imaging sampling moments when the same material to be sorted passes through the imaging area, the area and boundary position of the local bright candidate region are obtained respectively. Specifically, for each imaging sampling moment, the number of pixels contained in the local bright candidate region is determined as the area, and the pixel position adjacent to the non-bright response pixel in the local bright candidate region is determined as the boundary position. At the same time, using the material contour coverage position at the current imaging sampling moment as a reference, the position of the local bright candidate region in the material contour coverage position is converted into a relative region position. The relative region position is jointly determined by the contour boundary segment and the internal position of the local bright candidate region relative to the material contour coverage position.

[0022] After acquiring data at multiple imaging sampling times, the area corresponding to each imaging sampling time is arranged in chronological order to form an area change sequence; the boundary positions corresponding to each imaging sampling time are arranged in chronological order to form a boundary change sequence; and the relative area positions corresponding to each imaging sampling time are arranged in chronological order to form a relative area position sequence. Subsequently, the area change between adjacent imaging sampling times in the area change sequence is calculated, and the boundary position change between adjacent imaging sampling times in the boundary change sequence is calculated. An area change sample set is formed based on the area change between adjacent imaging sampling times in the local reflection sample data, and a preset area change range is determined from this sample set. Similarly, a boundary position change sample set is formed based on the boundary position change between adjacent imaging sampling times in the local reflection sample data, and a preset boundary position change range is determined from this sample set. When both the area change and boundary position change are within the preset range, the area change sequence and boundary position change sequence are deemed to satisfy a preset reflection morphology change condition.

[0023] When determining the relative region position sequence, the material contour coverage area is divided into multiple local contour position ranges according to the contour boundary arrangement order, and the local contour position range to which the relative region position belongs at each imaging sampling time is determined. When all relative region positions in the relative region position sequence belong to the same local contour position range, the relative region position sequence is determined to correspond to the same local contour position range within the material contour coverage area. Therefore, when the region area change sequence and region boundary change sequence satisfy the preset reflection morphology change condition, and the relative region position sequence corresponds to the same local contour position range within the material contour coverage area, the local bright candidate region is determined as the local bright region.

[0024] After identifying the locally highlighted areas, the non-highlighted normal areas are further identified within the material outline coverage area. Specifically, it is determined whether each pixel position within the material outline coverage area falls within the locally highlighted area; for pixel positions that do not fall within the locally highlighted area, it is further determined whether they have reached a saturated response state. When the brightness response value corresponding to a pixel position reaches a preset saturated response threshold, the pixel position is determined to have reached a saturated response state; wherein, the preset saturated response threshold is lower than the maximum response value of the imaging device. When a pixel position does not reach a saturated response state, and its brightness response value is within a preset effective response range, the pixel position is included in the non-highlighted normal area. The preset saturated response threshold is determined by subtracting a safety margin from the upper limit of the imaging device output, the preset low response threshold is determined by the upper limit of the background noise in the idle imaging data, and the preset effective response range is the brightness response range higher than the preset low response threshold and lower than the preset saturated response threshold.

[0025] After forming locally bright areas and non-bright normal areas, the change in brightness, area ratio, and overall brightness feedback change of the imaging area are calculated for the locally bright areas. Specifically, the change in brightness of the locally bright areas is determined by the difference between the brightness value of the locally bright areas in the current control cycle and the corresponding brightness value in the baseline control cycle; the area ratio of the locally bright areas is determined by the ratio between the area of ​​the locally bright areas and the area covered by the material outline; and the overall brightness feedback change of the imaging area is determined by the difference between the overall brightness value of the imaging area in the current control cycle and the overall brightness value of the imaging area in the baseline control cycle. The baseline control cycle is the control cycle collected and recorded under the baseline sorting state.

[0026] When the change in overall brightness feedback of the imaging area is greater than the preset positive threshold for brightness change, the direction of the change in overall brightness feedback of the imaging area is determined as the direction of brightness enhancement; when the change in overall brightness feedback of the imaging area is less than the preset negative threshold for brightness change, the direction of the change in overall brightness feedback of the imaging area is determined as the direction of brightness reduction; when the change in overall brightness feedback of the imaging area is within the preset stable range for brightness change, the direction of the change in overall brightness feedback of the imaging area is determined as the direction of brightness maintenance.

[0027] When generating the bright spot feedback traction, the non-high-brightness normal area is first divided into multiple normal feedback sub-regions. Specifically, according to the spatial partitioning order within the material outline coverage area, the non-high-brightness normal area is divided into multiple normal feedback sub-regions. Each normal feedback sub-region contains multiple pixel positions with brightness response values ​​within a preset effective response range. For each normal feedback sub-region, the difference between the sub-region brightness value in the current control cycle and the sub-region brightness value in the baseline control cycle is calculated to form the sub-region brightness change. Based on the brightness change samples of the non-high-brightness normal area under the baseline sorting state, the upper limit of positive fluctuation and the lower limit of negative fluctuation are statistically analyzed. The upper limit of positive fluctuation is determined as the preset positive threshold for brightness change, and the lower limit of negative fluctuation is determined as the preset negative threshold for brightness change. The range between the preset negative threshold for brightness change and the preset positive threshold for brightness change is determined as the preset stable range for brightness change. When the brightness change in a sub-region exceeds a preset positive brightness change threshold, the direction of the brightness change is defined as the brightness enhancement direction; when the brightness change is less than a preset negative brightness change threshold, the direction of the brightness change is defined as the brightness reduction direction; when the brightness change is within a preset stable brightness change range, the direction of the brightness change is defined as the brightness maintenance direction. The brightness change directions of each normal feedback sub-region are combined into a normal feedback direction set. The brightness enhancement direction, brightness reduction direction, and brightness maintenance direction are used to describe the trend of image brightness change; when subsequently generating illumination brightness control commands, the direction of increasing illumination brightness, maintaining illumination brightness, and decreasing illumination brightness are used to describe the control feedback direction.

[0028] Subsequently, the direction of brightness change in the locally highlighted area is determined based on the change in regional brightness. Specifically, when the change in regional brightness of the locally highlighted area is greater than a preset positive threshold, the direction of brightness change is determined as the brightness enhancement direction; when the change in regional brightness of the locally highlighted area is less than a preset negative threshold, the direction of brightness change is determined as the brightness reduction direction; and when the change in regional brightness of the locally highlighted area is within a preset stable range, the direction of brightness change is determined as the brightness maintenance direction. The consistency of the direction of brightness change in the locally highlighted area is compared with the direction of brightness change of each sub-region in the set of normal feedback directions. The number of normal feedback sub-regions that are inconsistent with the direction of brightness change in the locally highlighted area is counted, and the ratio between this number and the total number of normal feedback sub-regions is determined as the local direction deviation.

[0029] The calculation continues, determining the proportion of brightness contribution from the change in brightness of the local bright area to the overall brightness feedback change in the imaging area. Specifically, the absolute value of the change in brightness of the local bright area is proportional to the absolute value of the overall brightness feedback change in the imaging area to obtain the brightness contribution proportion. When the absolute value of the overall brightness feedback change in the imaging area is lower than a preset overall change threshold, the brightness contribution proportion for the current control cycle is marked as invalid, and the bright spot feedback traction amount for the current control cycle is also marked as invalid traction. When the bright spot feedback traction amount is marked as invalid, the current control cycle does not generate a bright spot feedback traction result based on that bright spot feedback traction amount. The preset overall change threshold is determined by the upper limit of background fluctuation in the distribution of the absolute value of the overall brightness feedback change in the imaging area under the baseline sorting state.

[0030] Subsequently, an area inverse vector is formed based on the inverse value of the area proportion of the local bright area. Specifically, the upper limit of the area proportion of the local reflective area in the local reflective sample data is statistically analyzed, and this upper limit is used as a preset area reference value; the area proportion of the local bright area is subtracted from the preset area reference value to obtain the area inverse vector; when the difference is less than zero, the area inverse vector is set to zero. The area inverse vector is used to participate in the calculation of the brightness spot feedback traction, so as to reflect the degree to which the local bright area participates in the overall brightness feedback change of the imaging area by occupying a local area.

[0031] After obtaining the local directional deviation, brightness contribution ratio, and area inverse vector, samples of the local directional deviation, brightness contribution ratio, and area inverse vector are extracted from the historical feedback samples, and the corresponding minimum and maximum values ​​are determined. The minimum value of each sample is used as the preset lower limit of the corresponding mapping, and the maximum value of each sample is used as the preset upper limit of the corresponding mapping. Subsequently, the local directional deviation, brightness contribution ratio, and area inverse vector are mapped to the same value range of 0 to 1, respectively, to obtain the local directional deviation mapping value, brightness contribution mapping value, and area inverse mapping value.

[0032] When determining the preset traction weights, multiple historical feedback samples are first acquired under the baseline sorting state. Each historical feedback sample includes the local directional deviation, brightness contribution ratio, area inverse vector, and a marker indicating whether bright spot-induced miscontrol occurred during the corresponding control cycle. Samples exhibiting bright spot-induced miscontrol are extracted from the historical feedback samples, and the average values ​​of the local directional deviation, brightness contribution ratio, and area inverse vector are calculated for each sample. The sum of these three average values ​​is used as the weight calculation benchmark, and the ratio of each average value to the weight calculation benchmark is determined as the preset traction weight corresponding to the local directional deviation mapping value, brightness contribution mapping value, and area inverse mapping value, respectively. Subsequently, the local directional deviation mapping value, brightness contribution mapping value, and area inverse mapping value are weighted and summed according to the preset traction weights to form the bright spot feedback traction amount. In this way, the bright spot feedback traction amount is jointly limited by the local directional deviation, brightness contribution ratio, and area inverse vector.

[0033] Step 2: Based on the brightness feedback traction amount, determine the corresponding brightness change direction of the local bright area and the overall brightness feedback change direction of the imaging area to form the brightness feedback traction result; Specifically, after completing step one and generating the brightness feedback traction amount corresponding to each local bright area, the brightness feedback traction amount corresponding to each local bright area is read within the current control cycle, and the local bright areas are sorted in descending order of brightness feedback traction amount. After sorting, the sorting sequence number of each local bright area in the sorting sequence is determined as the traction sorting position of that local bright area; where the higher the brightness feedback traction amount value, the earlier the corresponding traction sorting position. If there is only one local bright area in the current control cycle, that local bright area is determined as the first traction sorting position. Thus, the traction sorting position is used to represent the relative traction intensity of each local bright area on the overall brightness feedback change of the imaging area within the current control cycle.

[0034] After determining the traction sorting position, the direction of brightness change in the local bright area formed in step one and the direction of brightness feedback change in the overall imaging area are obtained, and their direction consistency is judged. Specifically, when both the direction of brightness change in the local bright area and the direction of brightness feedback change in the overall imaging area are brightness-increasing directions, their directions are determined to be consistent; when both the direction of brightness change in the local bright area and the direction of brightness feedback change in the overall imaging area are brightness-decreasing directions, their directions are determined to be consistent; when both the direction of brightness change in the local bright area and the direction of brightness feedback change in the overall imaging area are brightness-maintaining directions, their directions are determined to be consistent; when the direction of brightness change in the local bright area and the direction of brightness feedback change in the overall imaging area do not belong to the same direction among brightness-increasing, brightness-decreasing, and brightness-maintaining directions, their directions are determined to be inconsistent. Through this direction consistency judgment, it is determined whether the brightness change of the local bright area is in the same direction as the overall brightness feedback change of the imaging area.

[0035] After determining directional consistency, the traction sorting position is compared with the preset traction sorting range. The preset traction sorting range is determined by historical feedback data and erroneous control sample data under the baseline sorting state. Specifically, the traction quantity sorting results of the bright spots corresponding to the local bright areas in each control cycle are extracted from the historical feedback data, and the sorting position distribution of the corresponding local bright areas in the control cycle in which bright spots induce erroneous control is statistically analyzed. The sorting coverage ratio is determined based on the distribution of the number of erroneous control samples and equipment debugging data. The sorting coverage ratio is used to determine the preset traction sorting range from the sorting position distribution. The sorting range of the erroneous control samples corresponding to the sorting coverage ratio in the sorting position distribution is determined as the preset traction sorting range. When the traction sorting position of a certain local bright area is within the preset traction sorting range, the local bright area is determined to meet the traction sorting requirements.

[0036] When the direction of brightness change in a localized bright area is consistent with the direction of brightness feedback change in the overall imaging area, and the pulling ranking position of this localized bright area is within a preset pulling ranking range, this localized bright area is determined to be in a dominant pulling state and is designated as the dominant pulling area. When the direction of brightness change in a localized bright area is inconsistent with the direction of brightness feedback change in the overall imaging area, this localized bright area is not designated as the dominant pulling area. When the pulling ranking position of this localized bright area is not within the preset pulling ranking range, it is not designated as the dominant pulling area. Therefore, the dominant pulling area is not determined solely by the magnitude of the bright spot feedback pulling amount, but is jointly limited by the ranking and directional consistency of the bright spot feedback pulling amount.

[0037] After determining the dominant traction region, the brightness deduction amount corresponding to the dominant traction region is calculated. Specifically, the brightness change of the dominant traction region is read, and the area ratio of the dominant traction region is determined based on the ratio between the area of ​​the dominant traction region and the area covered by the material outline. Subsequently, the brightness change of the dominant traction region is multiplied by the area ratio of the dominant traction region to obtain the brightness deduction amount corresponding to the dominant traction region. The brightness deduction amount is the amount of data deducted from the overall brightness feedback change of the imaging area in the subsequent step three.

[0038] After determining the luminance subtraction amount corresponding to the dominant traction region, the region location, luminance change, luminance subtraction amount, traction sorting position, and dominant traction status of the dominant traction region are written into the bright spot feedback traction result. The region location is jointly represented by the set of pixel locations corresponding to the dominant traction region and the region boundary location; the luminance change amount is used to record the luminance change amplitude of the dominant traction region itself; the luminance subtraction amount is used for the bright spot subtraction processing in the subsequent step three; the traction sorting position is used to record the relative traction intensity of the dominant traction region within the current control cycle; and the dominant traction status indicates that the local bright area has been identified as the dominant traction region.

[0039] Step 3: Based on the brightness feedback traction results, perform brightness spot subtraction processing on the overall brightness feedback change of the imaging area to generate brightness feedback data with removed brightness spots, including: When the direction of the overall brightness feedback change of the imaging area is determined to be dominated by the local bright area based on the bright spot feedback traction result, the brightness deduction amount corresponding to the local bright area is determined based on the regional brightness change amount and area ratio of the local bright area, and the brightness deduction amount is deducted from the overall brightness feedback change amount of the imaging area to form bright spot removal brightness feedback data. When the direction of the overall brightness feedback change of the imaging area is not determined based on the brightness spot feedback traction result, the amount of overall brightness feedback change of the imaging area is used as the brightness feedback data for removing the brightness spot. Specifically, after completing step two and generating the bright spot feedback traction result, the dominant traction state corresponding to each local bright area in the bright spot feedback traction result is read. When there is no local bright area in the bright spot feedback traction result where the dominant traction state is established, the overall brightness feedback change direction of the imaging area where the local bright area is not determined within the current control cycle is determined, and the overall brightness feedback change of the imaging area is used as the brightness feedback data to be verified for bright spot removal.

[0040] When there is a locally bright area in the bright spot feedback pulling result where the dominant pulling state is established, the locally bright area where the dominant pulling state is established is determined as the dominant pulling area, and the area position and brightness deduction amount corresponding to each dominant pulling area are read from the bright spot feedback pulling result; wherein, the area position corresponding to the dominant pulling area is jointly represented by the set of pixel positions corresponding to the dominant pulling area and the area boundary position, and the brightness deduction amount is the data written into the bright spot feedback pulling result in step two.

[0041] When there is only one dominant traction region in the current control cycle, the brightness deduction amount corresponding to the dominant traction region is read from the brightness feedback traction result, and the brightness deduction amount is subtracted from the overall brightness feedback change of the imaging region to form the brightness feedback data to be verified for bright spot removal.

[0042] When multiple dominant traction regions exist within the current control cycle, for each dominant traction region, first calculate the sum of the luminance response values ​​corresponding to each pixel position within that region. Then, calculate the proportion of each pixel position's luminance response value in the sum of luminance response values. Based on this proportion, allocate the luminance deduction amount corresponding to the dominant traction region to each pixel position to obtain the pixel deduction amount for each pixel position. When the luminance response values ​​corresponding to each pixel position within the same dominant traction region are the same, distribute the luminance deduction amount corresponding to the dominant traction region equally among the pixel positions according to the number of pixels to obtain the pixel deduction amount for each pixel position.

[0043] After obtaining the pixel deduction values ​​corresponding to each pixel position within each dominant traction region, the pixel position sets of different dominant traction regions are compared and overlapped. When two dominant traction regions contain the same pixel position, the same pixel position is determined as the overlapping region position; pixel positions belonging to only one dominant traction region are determined as non-overlapping region positions. Thus, both overlapping and non-overlapping region positions are formed by comparing the pixel position sets corresponding to the dominant traction regions, avoiding the omission of locally overlapping pixels due to relying solely on region boundary positions.

[0044] After determining the locations of overlapping and non-overlapping regions, deduction amounts for both regions are generated. For overlapping regions, when the same pixel location belongs to multiple dominant traction regions, this embodiment employs a deduplication rule prioritizing traction ranking. Only the pixel deduction amount corresponding to the dominant traction region with the highest traction ranking is retained at that pixel location. The pixel deduction amounts for each overlapping region location are then summed to form the overlapping region deduction amount. For non-overlapping regions, the pixel deduction amounts corresponding to each non-overlapping region location are summed to form the non-overlapping region deduction amount. Therefore, the deduction amount corresponding to the same pixel location participates in the deduction only once in overlapping regions, avoiding duplicate deductions when multiple dominant traction regions overlap.

[0045] After obtaining the overlapping area deduction and the non-overlapping area deduction, the overlapping area deduction and the non-overlapping area deduction are summed to form the deduplication deduction; then, the deduplication deduction is subtracted from the overall brightness feedback change of the imaging area to form the deduction residue.

[0046] In this embodiment, the deducted residual amount is the data formed by subtracting the deduplication amount from the overall brightness feedback change of the imaging area. It is used to compare with the residual verification range, and after the comparison, the verified brightness feedback data or the corrected brightness feedback data is formed.

[0047] After deducting the residual amount, the brightness change of the non-high-brightness normal area is calculated, and a residual verification range is formed based on the brightness change of the non-high-brightness normal area and a preset brightness deviation range. Specifically, the brightness change of the non-high-brightness normal area is determined by the difference between the brightness value of the non-high-brightness normal area in the current control cycle and the corresponding brightness value in the reference control cycle; the preset brightness deviation range is determined by the fluctuation data of the brightness change of the non-high-brightness normal area under the reference sorting state. The preset brightness deviation range includes positive deviation values ​​and negative deviation values, where the positive deviation value serves as the upper boundary of the preset brightness deviation range, and the negative deviation value serves as the lower boundary of the preset brightness deviation range; taking the brightness change of the non-high-brightness normal area as the center, the brightness change of the non-high-brightness normal area is added to the positive deviation value to form the upper boundary of the residual verification range, and the brightness change of the non-high-brightness normal area is added to the negative deviation value to form the lower boundary of the residual verification range, thus forming the residual verification range.

[0048] After establishing the residual verification range, the subtracted residual amount is compared with the residual verification range. When the subtracted residual amount is within the residual verification range, it is determined as the verified brightness feedback data after removing bright spots. When the subtracted residual amount is lower than the lower boundary value of the residual verification range, the lower boundary value of the residual verification range is determined as the boundary value closest to the subtracted residual amount, and the subtracted residual amount is limited to this lower boundary value to form the corrected brightness feedback data after removing bright spots. When the subtracted residual amount is higher than the upper boundary value of the residual verification range, the upper boundary value of the residual verification range is determined as the boundary value closest to the subtracted residual amount, and the subtracted residual amount is limited to this upper boundary value to form the corrected brightness feedback data after removing bright spots.

[0049] In this embodiment, when verified brightness feedback data for removing bright spots is generated, it is used as the brightness feedback data for step four; when corrected brightness feedback data for removing bright spots is generated, it is used as the brightness feedback data for step four; when no dominant traction region is formed within the current control cycle, the overall brightness feedback change of the imaging region is used as the brightness feedback data for step four. Therefore, the brightness feedback data output from step three to step four has a unified data name and a clear data source.

[0050] Step 4: Based on the brightness feedback data of the bright spots, the regional brightness change of the local bright areas, and the light attenuation feedback data of the lighting units, determine the brightness feedback direction corresponding to the non-bright normal areas, the brightness reduction feedback direction of the local bright areas, and the light attenuation compensation feedback direction of the lighting units, respectively, and compare the directional relationships to form the feedback direction splitting result. Specifically, after completing step three and generating the brightness feedback data for step four, the brightness feedback direction corresponding to the non-high-brightness normal area is first determined based on the brightness feedback data. Specifically, the brightness feedback data is compared with a preset brightness feedback range corresponding to the baseline brightness state. When the brightness feedback data is below the lower boundary of the preset brightness feedback range, the brightness feedback direction corresponding to the non-high-brightness normal area is determined to be the direction of increasing illumination brightness. When the brightness feedback data is within the preset brightness feedback range, the brightness feedback direction corresponding to the non-high-brightness normal area is determined to be the direction of maintaining illumination brightness. When the brightness feedback data is above the upper boundary of the preset brightness feedback range, the brightness feedback direction corresponding to the non-high-brightness normal area is determined to be the direction of decreasing illumination brightness. The preset brightness feedback range is determined by the fluctuation range of the brightness feedback data of the non-high-brightness normal area under the baseline sorting state.

[0051] After determining the brightness feedback direction corresponding to the non-high-brightness normal area, the brightness reduction feedback direction for the local high-brightness area is determined based on the regional brightness change of the local high-brightness area. Specifically, the regional brightness change of the local high-brightness area is compared with a preset brightness change condition. The preset brightness change condition is determined by the distribution of brightness change in the reflection area in the local reflection sample data under the baseline sorting state. Specifically, a threshold value for regional brightness change used to distinguish the local reflection high-brightness state is extracted from the distribution of brightness change in the reflection area, and this threshold value is used as the preset brightness change condition. When the change in brightness of a locally bright area reaches the preset brightness change condition, it is determined that the locally bright area has a brightness reduction feedback requirement caused by the local brightness, and the brightness reduction feedback direction of the locally bright area is determined as the direction of reducing lighting brightness; when the change in brightness of a locally bright area does not reach the preset brightness change condition, it is determined that the locally bright area has not formed a brightness reduction feedback direction, and the locally bright area is recorded as a locally bright area that has not triggered brightness reduction feedback; locally bright areas that have not triggered brightness reduction feedback do not participate in the direction consistency comparison between the brightness reduction feedback direction of the locally bright area and the normal compensation reference direction.

[0052] After determining the brightness reduction feedback direction for local high-brightness areas, the light attenuation compensation feedback direction for the lighting unit is determined based on the light attenuation feedback data of the lighting unit. Specifically, the current light output intensity, reference light output intensity, and light attenuation amount are read from the light attenuation feedback data of the lighting unit; the preset light attenuation amount condition is determined by the light output detection data of the lighting unit in the reference sorting state, specifically based on the attenuation distribution of the light output intensity relative to the reference light output intensity in the reference sorting state; the preset stable light output range is formed by the fluctuation range of the light output intensity in the reference sorting state; the preset positive fluctuation upper limit is determined by the fluctuation data of the light output intensity being higher than the reference light output intensity in the reference sorting state, and the preset upper limit of light output is jointly determined by the reference light output intensity and the preset positive fluctuation upper limit. If the current light output intensity is lower than the reference light output intensity and the light attenuation amount reaches the preset light attenuation amount condition, the light attenuation compensation feedback direction of the lighting unit is determined to be the direction of increasing lighting brightness; if the current light output intensity is within the preset stable light output range, the light attenuation compensation feedback direction of the lighting unit is determined to be the direction of maintaining lighting brightness; if the current light output intensity is higher than the preset upper limit of light output, the light attenuation compensation feedback direction of the lighting unit is determined to be the direction of decreasing lighting brightness.

[0053] After obtaining the brightness feedback direction corresponding to the non-high-brightness normal area, the brightness reduction feedback direction for the locally high-brightness area, and the light attenuation compensation feedback direction for the lighting unit, the brightness feedback direction corresponding to the non-high-brightness normal area is first compared with the light attenuation compensation feedback direction for the lighting unit to form a normal compensation direction correspondence result. Specifically, when both the brightness feedback direction corresponding to the non-high-brightness normal area and the light attenuation compensation feedback direction for the lighting unit are directions that increase lighting brightness, they are determined to be consistent; when both are directions that maintain lighting brightness, they are determined to be consistent; when both are directions that decrease lighting brightness, they are determined to be consistent; when they do not belong to the same direction among the directions of increasing lighting brightness, maintaining lighting brightness, and decreasing lighting brightness, they are determined to be inconsistent. The normal compensation direction correspondence result includes a consistent direction field, an inconsistent direction field, and a feedback direction field corresponding to the consistent direction.

[0054] When the result of the normal compensation direction determines that the brightness feedback direction corresponding to the non-high-brightness normal area is consistent with the light attenuation compensation feedback direction of the illumination unit, the feedback direction corresponding to the consistency between the two is extracted, and this feedback direction is determined as the normal compensation reference direction. The normal compensation reference direction can be one of the following: increasing illumination brightness, maintaining illumination brightness, or decreasing illumination brightness. Therefore, the normal compensation reference direction is not solely derived from the brightness feedback of the imaging area, nor solely from the light attenuation feedback of the illumination unit, but is jointly formed when the brightness feedback direction corresponding to the non-high-brightness normal area and the light attenuation compensation feedback direction of the illumination unit are consistent. This reference direction is used to subsequently determine whether the brightness reduction feedback direction of the local high-brightness area deviates from the normal brightness feedback requirements and light attenuation compensation requirements.

[0055] After establishing the normal compensation reference direction, the brightness reduction feedback direction of the local high-brightness area is compared with the normal compensation reference direction for directional consistency. When both the local high-brightness area's brightness reduction feedback direction and the normal compensation reference direction are in the direction of reducing illumination brightness, the local high-brightness area's brightness reduction feedback direction is determined to be consistent with the normal compensation reference direction. When the local high-brightness area's brightness reduction feedback direction is in the direction of reducing illumination brightness, while the normal compensation reference direction is either in the direction of maintaining illumination brightness or in the direction of increasing illumination brightness, the local high-brightness area's brightness reduction feedback direction is determined to be inconsistent with the normal compensation reference direction. This comparison determines whether the brightness reduction feedback triggered by the local high-brightness area is separated from the feedback direction jointly pointed to by the non-high-brightness normal area and the illumination unit's light attenuation compensation.

[0056] When the brightness reduction feedback direction of a locally highlighted area is inconsistent with the normal compensation reference direction, this locally highlighted area is identified as the directional split source region, and a directional split marker corresponding to the directional split source region and the brightness reduction feedback direction of the locally highlighted area is generated. Specifically, the region location of the directional split source region, the brightness reduction feedback direction of the locally highlighted area, the normal compensation reference direction, the control cycle identifier, and the directional split status are written into the directional split marker. The directional split status indicates that the brightness reduction feedback direction of the locally highlighted area is inconsistent with the normal compensation reference direction. The region location of the directional split source region is jointly represented by the set of pixel positions corresponding to the locally highlighted area and the region boundary position.

[0057] After forming the directional splitting markers, the brightness feedback direction corresponding to the non-high-brightness normal area, the light attenuation compensation feedback direction of the illumination unit, the brightness reduction feedback direction of the local high-brightness area, the location of the area corresponding to the directional splitting source area, and the directional splitting markers are written into the feedback directional splitting result. The feedback directional splitting result also records the control cycle identifier, so that the feedback directional splitting result can correspond to the brightness feedback data of the bright spot removal, the light attenuation feedback data of the illumination unit, and the local high-brightness area within the same control cycle. Thus, the feedback directional splitting result not only records the directional relationship of the three types of feedback directions, but also records the directional splitting source area and the directional splitting markers, providing a data basis for the subsequent step five to form the bright spot reduction feedback stripping result.

[0058] In this embodiment, when the normal compensation direction determines that the brightness feedback direction corresponding to the non-high-brightness normal area is inconsistent with the light attenuation compensation feedback direction of the lighting unit, the feedback direction splitting result is not generated according to the normal compensation reference direction formation path of this embodiment; the brightness feedback direction corresponding to the non-high-brightness normal area, the light attenuation compensation feedback direction of the lighting unit, the brightness reduction feedback direction of the local high-brightness area and the control cycle identifier within the control cycle are saved, and the saved data is not used as the basis for the formation of the bright spot brightness reduction feedback stripping result in this embodiment.

[0059] Step 5: Based on the feedback direction splitting results, generate bright spot reduction feedback stripping results; according to the bright spot reduction feedback stripping results, do not write the local high brightness area reduction feedback direction into the global lighting brightness control basis set, and write the bright spot reduction brightness feedback data and lighting unit light attenuation feedback data into the global lighting brightness control basis set to generate lighting brightness control instructions; Specifically, after completing step four and generating the feedback direction splitting result, the region location corresponding to the direction splitting source region, the local bright area reduction feedback direction, the direction splitting marker, and the control cycle identifier are read from the feedback direction splitting result. When the feedback direction splitting result contains a direction splitting marker, and the direction splitting status indication in the direction splitting marker is inconsistent with the local bright area reduction feedback direction and the normal compensation reference direction, the local bright area reduction feedback direction corresponding to the direction splitting source region is determined as the feedback direction to be stripped, and a bright spot reduction feedback stripping result is generated.

[0060] The bright spot reduction feedback stripping result includes the stripping object, stripping direction, stripping source location, stripping control cycle, and stripping status. The stripping object is the local bright area corresponding to the directional split source region; the stripping direction is the bright spot reduction feedback direction; the stripping source location is the region location corresponding to the directional split source region; the stripping control cycle is the control cycle identifier recorded in the feedback directional split result; and the stripping status records that the bright spot reduction feedback direction does not participate in the formation of the global illumination brightness control basis set. Therefore, the stripping object, stripping direction, stripping source location, stripping control cycle, and stripping status are written into the bright spot reduction feedback stripping result.

[0061] After generating the bright spot reduction feedback stripping result, a global lighting brightness control basis set corresponding to the current control cycle is established. Specifically, the bright spot removal brightness feedback data output in step three and the lighting unit light attenuation feedback data obtained in step one are read and written into the global lighting brightness control basis set. Simultaneously, the brightness feedback direction corresponding to the non-high-brightness normal area, the lighting unit light attenuation compensation feedback direction, and the control cycle identifier are recorded in the global lighting brightness control basis set. When the stripping status in the bright spot reduction feedback stripping result indicates that the local high-brightness area reduction feedback direction needs to be stripped, the local high-brightness area reduction feedback direction is not written into the global lighting brightness control basis set.

[0062] After forming a global lighting brightness control basis set, lighting brightness control instructions are generated based on this set. Specifically, a basic control direction is first formed based on the positional relationship between the bright spot brightness feedback data and the preset brightness feedback range: when the bright spot brightness feedback data is lower than the lower boundary value of the preset brightness feedback range, the basic control direction is determined to increase the lighting brightness; when the bright spot brightness feedback data is within the preset brightness feedback range, the basic control direction is determined to maintain the lighting brightness; when the bright spot brightness feedback data is higher than the upper boundary value of the preset brightness feedback range, the basic control direction is determined to decrease the lighting brightness.

[0063] After determining the basic control direction, a basic brightness control amount corresponding to the basic control direction is generated based on the deviation of the brightness feedback data of the de-spotted area from the boundary value of the preset brightness feedback range. Specifically, when the brightness feedback data of the de-spotted area is lower than the lower boundary value of the preset brightness feedback range, the difference between the lower boundary value of the preset brightness feedback range and the brightness feedback data of the de-spotted area is determined as the brightness low deviation, and a basic brightness control amount corresponding to the direction of increasing lighting brightness is formed based on the brightness low deviation; when the brightness feedback data of the de-spotted area is higher than the upper boundary value of the preset brightness feedback range, the difference between the brightness feedback data of the de-spotted area and the upper boundary value of the preset brightness feedback range is determined as the brightness high deviation, and a basic brightness control amount corresponding to the direction of decreasing lighting brightness is formed based on the brightness high deviation; when the brightness feedback data of the de-spotted area is within the preset brightness feedback range, the basic brightness control amount is determined to be zero.

[0064] Subsequently, a light attenuation compensation control amount is generated based on the light attenuation feedback data of the illumination unit. Specifically, the light attenuation amount is read from the light attenuation feedback data of the illumination unit, and converted into a light attenuation compensation control amount according to the correspondence between the preset light attenuation amount and the illumination brightness control amount; wherein, the light attenuation compensation control amount and the basic brightness control amount adopt the same brightness control scale. The correspondence between the preset light attenuation amount and the illumination brightness control amount is formed by the correspondence data between the change in light output intensity of the illumination unit and the change in brightness feedback of the imaging area under the reference sorting state.

[0065] After both the basic brightness control value and the light attenuation compensation control value are converted to the same brightness control scale, a lighting brightness control command is generated based on the basic control direction and the light attenuation compensation feedback direction of the lighting unit. When the basic control direction is consistent with the light attenuation compensation feedback direction of the lighting unit, the basic brightness control value and the light attenuation compensation control value are summed to obtain the lighting brightness control value in the corresponding direction, and a lighting brightness control command in the corresponding direction is generated based on the lighting brightness control value. When the light attenuation compensation feedback direction of the lighting unit is to maintain the lighting brightness direction, a lighting brightness control command is generated according to the basic control direction and the basic brightness control value. When the basic control direction is to maintain the lighting brightness direction and the light attenuation compensation feedback direction of the lighting unit is not to maintain the lighting brightness direction, a lighting brightness control command is generated according to the light attenuation compensation feedback direction of the lighting unit and the light attenuation compensation control value.

[0066] When the basic control direction is opposite to the light attenuation compensation feedback direction of the lighting unit, the basic brightness control amount is compared with the light attenuation compensation control amount. When the basic brightness control amount is greater than the light attenuation compensation control amount, a lighting brightness control command is generated according to the basic control direction and the difference between the basic brightness control amount and the light attenuation compensation control amount. When the basic brightness control amount is less than the light attenuation compensation control amount, a lighting brightness control command is generated according to the light attenuation compensation feedback direction of the lighting unit and the difference between the light attenuation compensation control amount and the basic brightness control amount. When the basic brightness control amount is equal to the light attenuation compensation control amount, a lighting brightness control command that maintains the lighting brightness direction is generated.

[0067] Example 2

[0068] This embodiment, based on the bright spot reduction feedback stripping result formed by the feedback direction splitting result in Embodiment 1, further introduces the local transient bright spot determination result and the imaging region representative migration determination result to limit the formation conditions of the bright spot reduction feedback stripping result. Specifically, after completing step four of Embodiment 1 and forming the feedback direction splitting result, the region position corresponding to the direction splitting source region, the local bright area reduction feedback direction, the normal compensation reference direction, the control cycle identifier, and the direction splitting marker in the feedback direction splitting result are read, and the local bright area corresponding to the direction splitting source region is determined as the local bright area to be determined.

[0069] Within a continuous control cycle, the location, boundary position, and area of ​​the local bright area to be judged are acquired in each control cycle. The number of continuous control cycles is determined by the number of control cycles required for the material to be sorted to pass through the imaging area under the baseline sorting state. For two adjacent control cycles, the change in the center position of the local bright area to be judged is calculated to establish a positional migration relationship; the overlapping area of ​​the local bright area to be judged in two adjacent control cycles is calculated, and the area of ​​the local bright area to be judged in the two control cycles is read respectively. The minimum of the two area values ​​is determined as the overlapping reference area, and the ratio between the overlapping area and the overlapping reference area is calculated to establish a region overlap relationship; the number of consecutive control cycles in which the local bright area to be judged is identified as a local bright area is counted to form the number of continuous control cycles.

[0070] After establishing the position migration relationship, region overlap relationship, and number of continuous control cycles, the position migration relationship is compared with the transport direction of the material to be sorted within the imaging area. The region overlap relationship is compared with a preset overlap range, and the number of continuous control cycles is compared with a preset range of continuous control cycles. When the position migration relationship is consistent with the transport direction of the material to be sorted within the imaging area, the region overlap relationship is within the preset overlap range, and the number of continuous control cycles is within the preset range of continuous control cycles, a local transient bright spot determination result is formed to indicate that the local bright area to be determined belongs to a local transient bright spot that moves with the material. The preset overlap range and the preset range of continuous control cycles are determined by the region overlap data and continuous cycle data of the local bright area moving with the material in the local reflection sample data under the baseline sorting state.

[0071] After the local bright area to be determined exits the imaging area, it enters a preset verification control cycle range. The preset verification control cycle range is determined by the number of control cycles required for the brightness feedback direction of the non-bright normal area to recover after the local bright area exits the imaging area under the baseline sorting state. Within the preset verification control cycle range, the brightness feedback direction corresponding to the non-bright normal area in each verification control cycle is read one by one, and its consistency with the brightness reduction feedback direction of the local bright area is compared; when the brightness feedback direction corresponding to the non-bright normal area in each verification control cycle is not consistent with the brightness reduction feedback direction of the local bright area, a representative migration judgment result of the imaging area is formed to indicate that the brightness feedback direction corresponding to the non-bright normal area does not continue the brightness reduction feedback direction of the local bright area.

[0072] Subsequently, the results of the local transient bright spot determination, the representative migration determination of the imaging area, and the feedback direction splitting result are jointly judged. When the local transient bright spot determination result determines that the local bright area to be determined belongs to a local transient bright spot that moves with the material, the representative migration determination of the imaging area determines that the brightness feedback direction corresponding to the non-bright normal area does not continue the brightness reduction feedback direction of the local bright area, and the feedback direction splitting result determines that the brightness reduction feedback direction of the local bright area is split from the brightness feedback direction corresponding to the non-bright normal area and the light attenuation compensation feedback direction of the illumination unit, the brightness reduction feedback direction of the local bright area corresponding to the local bright area to be determined is determined as the feedback direction that needs to be stripped, and a bright spot brightness reduction feedback stripping result is formed.

[0073] The brightness reduction feedback stripping result includes the stripping object, stripping direction, stripping source location, stripping control cycle, and stripping status. The stripping object is the local bright area corresponding to the directional split source region; the stripping direction is the brightness reduction feedback direction of the local bright area; the stripping source location is the region location corresponding to the directional split source region; the stripping control cycle is the control cycle identifier recorded in the feedback directional split result; and the stripping status is used to record that the brightness reduction feedback direction of the local bright area does not participate in the formation of the global lighting brightness control basis set.

[0074] After generating the bright spot reduction feedback stripping result, the bright spot removal brightness feedback data output in step three of Example 1 and the illumination unit light attenuation feedback data obtained in step one are read, and the bright spot removal brightness feedback data and illumination unit light attenuation feedback data are written into the global illumination brightness control basis set. Simultaneously, the brightness feedback direction corresponding to the non-high-brightness normal area, the illumination unit light attenuation compensation feedback direction, and the control cycle identifier are synchronously recorded in the global illumination brightness control basis set. When the stripping status in the bright spot reduction feedback stripping result indicates that the local high-brightness area reduction feedback direction needs to be stripped, the local high-brightness area reduction feedback direction is not written into the global illumination brightness control basis set, and an illumination brightness control command is generated according to the method disclosed in step five of Example 1.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback, characterized in that, include: Step 1: Obtain imaging area feedback data and illumination unit light attenuation feedback data. Separate local bright areas and non-bright normal areas from the imaging area feedback data. Based on the brightness change, area ratio of the local bright areas and the overall brightness feedback change of the imaging area, form the bright spot feedback traction amount. Step 2: Based on the brightness feedback traction amount, determine the corresponding brightness change direction of the local bright area and the overall brightness feedback change direction of the imaging area to form the brightness feedback traction result; Step 3: Based on the bright spot feedback traction results, perform bright spot subtraction processing on the overall brightness feedback change of the imaging area to form bright spot-removed brightness feedback data; Step 4: Based on the brightness feedback data of the bright spots, the regional brightness change of the local bright areas, and the light attenuation feedback data of the lighting units, determine the brightness feedback direction corresponding to the non-bright normal areas, the brightness reduction feedback direction of the local bright areas, and the light attenuation compensation feedback direction of the lighting units, respectively, and compare the directional relationships to form the feedback direction splitting result. Step 5: Generate bright spot reduction feedback stripping results based on feedback direction splitting results; according to the bright spot reduction feedback stripping results, do not write the local high brightness area reduction feedback direction into the global lighting brightness control basis set, and write the bright spot reduction brightness feedback data and lighting unit light attenuation feedback data into the global lighting brightness control basis set to generate lighting brightness control instructions.

2. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step 3 generates brightness feedback data for removing bright spots, including: When the direction of the overall brightness feedback change of the imaging area is determined to be dominated by the local bright area based on the bright spot feedback traction result, the brightness deduction amount corresponding to the local bright area is determined based on the regional brightness change amount and area ratio of the local bright area, and the brightness deduction amount is deducted from the overall brightness feedback change amount of the imaging area to form bright spot removal brightness feedback data. When the direction of the overall brightness feedback change of the imaging area is not determined based on the brightness spot feedback traction result, the amount of overall brightness feedback change of the imaging area is used as the brightness feedback data for removing the brightness spot.

3. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step five involves generating bright spot reduction feedback stripping results based on the feedback direction splitting results, including: Based on the positional migration relationship, regional overlap relationship, and number of continuous control cycles of the local bright areas, a local transient bright spot determination result is formed. Within a preset verification control period after a locally bright area exits the imaging area, it is determined whether the brightness feedback direction of the non-bright normal area continues the brightness reduction feedback direction of the locally bright area, thus forming a representative migration judgment result of the imaging area. When the local transient bright spot determination result determines that the local high-brightness area belongs to the local transient bright spot that moves with the material, the imaging area representative migration determination result determines that the brightness feedback direction corresponding to the non-high-brightness normal area does not continue the brightness reduction feedback direction of the local high-brightness area, and the feedback direction split result determines that the brightness reduction feedback direction of the local high-brightness area is split from the brightness feedback direction corresponding to the non-high-brightness normal area and the light attenuation compensation feedback direction of the illumination unit, a bright spot brightness reduction feedback stripping result is formed.

4. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step one, which involves separating locally highlighted areas from non-high-brightness normal areas in the imaging area feedback data, includes: In the imaging area feedback data, the pixel position and brightness response value of the material to be sorted are extracted based on the difference in brightness response between the pixel position and the background area. The contour boundary position is determined based on the pixel position of the material to be sorted, and the contour boundary position is used to enclose the material contour coverage position. Within the area covered by the material outline, multiple pixel positions are selected as high-brightness response positions in descending order of the brightness response values ​​corresponding to each pixel position, and the high-brightness response positions that satisfy the pixel adjacency relationship are connected to form a local high-brightness candidate area. During multiple imaging sampling moments when the same material to be sorted passes through the imaging area, the area and boundary position of the local bright candidate region are obtained respectively. The position of the local bright candidate region in the material outline coverage position is converted into the relative region position. According to the order of multiple imaging sampling moments, the region area change sequence, the region boundary change sequence, and the relative region position sequence are formed. When the region area change sequence and the region boundary change sequence meet the preset reflection shape change conditions, and the relative region position sequence corresponds to the same local contour position range within the material contour coverage position, the local highlight candidate region is determined as the local highlight region. The set of pixel locations within the material outline coverage area that do not fall into the local bright area, do not reach the saturation response state, and whose brightness response value is within the preset effective response range are determined as the non-bright normal area.

5. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step one generates a bright spot feedback traction amount, including: The non-highlight normal area is divided into multiple normal feedback sub-regions, and the brightness change direction of each sub-region is calculated to form a set of normal feedback directions. The direction of brightness change in the local bright area is determined based on the change in brightness of the local bright area. The consistency of the direction of brightness change in the local bright area with the direction of brightness change in each sub-region in the set of normal feedback directions is compared. The proportion of normal feedback sub-regions that are inconsistent with the direction of brightness change in the local bright area is counted to form the local direction deviation. Calculate the proportion of the brightness change of the local bright area in the overall brightness feedback change of the imaging area, and form an area inverse vector based on the inverse value of the area proportion of the local bright area. By mapping the local directional deviation, the proportion of brightness contribution, and the area inverse vector to the same value range, we obtain the local directional deviation mapping value, the brightness contribution mapping value, and the area inverse mapping value. The local directional deviation mapping value, brightness contribution mapping value, and area reverse mapping value are weighted and summed according to the preset traction weights to form the bright spot feedback traction amount.

6. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step two, which generates the bright spot feedback traction result, includes: The traction order of the local bright areas is determined according to the order of the bright spot feedback traction amount from high to low. When the direction of brightness change in a local bright area is consistent with the direction of brightness feedback change in the overall imaging area, and the traction sorting position is within the preset traction sorting range, the local bright area is determined to be in the dominant traction state, and the local bright area is determined to be the dominant traction area. The brightness deduction amount corresponding to the dominant traction area is determined based on the regional brightness change and area ratio of the dominant traction area. The region location of the dominant traction area, the amount of regional brightness change, the brightness deduction, the traction sorting position, and the dominant traction status are written into the bright spot feedback traction results.

7. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 6, characterized in that, After generating the brightness feedback data for removing bright spots in step three, the brightness feedback data is corrected, including: Read the region location and brightness deduction amount corresponding to the dominant traction region from the bright spot feedback traction results, and compare the overlapping regions corresponding to each dominant traction region to determine the overlapping and non-overlapping regions. The brightness deduction amounts corresponding to the overlapping area positions are merged according to the rule of retaining the same pixel position once to obtain the overlapping area deduction amount, and the brightness deduction amounts corresponding to the non-overlapping area positions are determined as the non-overlapping area deduction amount. The deduction amounts for overlapping areas and non-overlapping areas are summed to form the deduplication amount; The deduplication amount is subtracted from the overall brightness feedback change in the imaging area to form the deduplication residual amount; Calculate the brightness change of the non-highlight normal area, and form a residual verification range based on the brightness change of the non-highlight normal area and the preset brightness deviation range; The residual amount is compared with the residual verification range. When the residual amount is within the residual verification range, the residual amount is determined as the verified brightness feedback data after removing bright spots. When the deducted residual amount is not within the residual verification range, the deducted residual amount is limited and corrected according to the boundary value closest to the deducted residual amount in the residual verification range, thus forming the corrected brightness feedback data for removing bright spots.

8. The adaptive control method for imaging parameters of a photoelectric sorting system based on real-time feedback according to claim 1, characterized in that, Step four generates the feedback direction splitting result, including: The brightness feedback direction corresponding to the non-high-brightness normal area is compared with the light attenuation compensation feedback direction of the lighting unit to form the result corresponding to the normal compensation direction. When the result of the normal compensation direction determines that the brightness feedback direction corresponding to the non-high brightness normal area is consistent with the light attenuation compensation feedback direction of the lighting unit, the feedback direction corresponding to the consistency between the two is extracted, and the feedback direction is determined as the normal compensation reference direction. The brightness reduction feedback direction of the local bright area is compared with the normal compensation reference direction. When the brightness reduction feedback direction of the local bright area is inconsistent with the normal compensation reference direction, the local bright area is identified as the direction split source area, and a direction split mark corresponding to the direction split source area and the brightness reduction feedback direction of the local bright area is generated. Write the brightness feedback direction corresponding to the non-high-brightness normal area, the light attenuation compensation feedback direction of the illumination unit, the brightness reduction feedback direction of the local high-brightness area, the area location corresponding to the direction split source area, and the direction split mark into the feedback direction split result.

Images