CN-122007058-A - Rubber block sorting method and system based on thermal imaging and near infrared

CN122007058ACN 122007058 ACN122007058 ACN 122007058ACN-122007058-A

Abstract

The invention relates to the technical field of rubber block sorting, and discloses a rubber block sorting method and system based on thermal imaging and near infrared, wherein the method comprises the steps of conveying rubber blocks to a separation, detection and sorting station; the method comprises the steps of separating at a separating station, collecting thermal imaging images at a detecting station to obtain outlines, temperature distribution and coordinates of candidate abnormal areas, controlling a near infrared device to sample the candidate abnormal areas at fixed points to obtain absorbance of characteristic wavelengths of moisture, performing temperature compensation on the absorbance by using corresponding surface temperature, then performing sorting judgment, and generating instructions to control the action of an executing mechanism according to judgment results, positions and speeds. According to the invention, near infrared fixed point sampling is guided through thermal imaging positioning and temperature compensation is combined, so that the problem of insufficient sorting synchronization precision in the prior art is solved, and the accuracy, pertinence and stability of rubber block sorting are improved.

Inventors

YU ZHIQIANG
CHEN LEI
HUANG YU
ZHANG KE

Assignees

北京理工大学长三角研究院(嘉兴)

Dates

Publication Date: 20260512
Application Date: 20260327

Claims (10)

1. A thermal imaging and near infrared based rubber block sorting method, comprising: sequentially conveying rubber blocks to be sorted along the conveying direction to pass through a separation station, a detection station and a sorting station, wherein the separation station is used for separating adhered or stacked rubber blocks and outputting single-layer dispersed rubber blocks; Acquiring thermal imaging images of the rubber blocks at the detection station by using a thermal imaging device, performing image processing on the thermal imaging images to obtain contour information and surface temperature distribution information of each rubber block, and determining position coordinates of candidate abnormal rubber blocks and candidate abnormal areas thereof based on the surface temperature distribution information; According to the position coordinates of the candidate abnormal region, controlling a near infrared spectrum device to sample the candidate abnormal region at a near infrared detection position at a fixed point to obtain near infrared absorbance data at least one moisture characteristic wavelength; Acquiring the surface temperature of the candidate abnormal region corresponding to the fixed point sampling, taking the surface temperature as compensation input, and performing temperature compensation on the near infrared absorbance data to obtain a temperature correction near infrared characteristic parameter; Sorting judgment is carried out based on the temperature correction near infrared characteristic parameters, and sorting judgment results are output; and generating a sorting instruction according to the sorting judgment result and the position and the conveying speed of the candidate abnormal rubber blocks in the conveying direction, and controlling a sorting executing mechanism to execute a splitting or rejecting action on the corresponding rubber blocks at the sorting station.
2. The thermal imaging and near infrared based rubber block sorting method according to claim 1, characterized in that when the separation process is performed on the adhered or stacked rubber blocks at the separation station, it comprises: The separation station is divided into a feeding section, a stripping section and an accelerating section which are sequentially connected, the stripping section is provided with a conveying speed different from that of the feeding section and applies stripping action to the rubber block, the stripping section is provided with vibration action to the rubber block, the tail end of the stripping section is used for detecting the lamination state of the rubber block, and parameters of the conveying speed and the vibration action are adjusted according to the lamination state.
3. The thermal imaging and near infrared based rubber block sorting method according to claim 2, wherein when performing image processing on the thermal imaging image to obtain profile information and surface temperature distribution information of each rubber block, comprising: Performing radiation correction processing and noise suppression processing on the thermal imaging image, performing background separation to obtain a rubber block area, performing connected domain analysis on the rubber block area, removing a non-target area, performing boundary extraction on a target connected domain to obtain the contour information, extracting pixel temperature in an area limited by the contour information to obtain the surface temperature distribution information, and establishing a corresponding mapping between pixel coordinates and conveying coordinates.
4. A thermal imaging and near infrared based rubber block sorting method according to claim 3, characterized in that when determining the position coordinates of the candidate abnormal rubber block and its candidate abnormal region based on the surface temperature distribution information, it comprises: the method includes the steps of defining contour information of a rubber block, calculating surface temperature statistics in a region defined by the contour information, identifying a temperature deviation region having deviation relative to the surface temperature statistics in the surface temperature distribution information, determining the temperature deviation region as a candidate abnormal region, performing region screening on the candidate abnormal region and determining the geometric position of the candidate abnormal region, marking the rubber block containing the candidate abnormal region as a candidate abnormal rubber block, and outputting position coordinates of the candidate abnormal region.
5. The thermal imaging and near infrared based rubber block sorting method of claim 4, wherein when said detection station acquires a thermal imaging image of said rubber block with a thermal imaging device, comprising: The method comprises the steps of collecting at least two thermal imaging images of the same rubber block in the conveying direction, obtaining temperature drop characteristics according to collected surface temperature distribution information, obtaining quality information of the rubber block and correcting the temperature drop characteristics, and collecting the quality information through a dynamic weighing device.
6. The method for sorting rubber blocks based on thermal imaging and near infrared according to claim 5, wherein controlling the near infrared spectroscopy device to sample the candidate abnormal region at a near infrared detection position according to the position coordinates of the candidate abnormal region comprises: the method comprises the steps of obtaining conveying speed information and position information, converting position coordinates of the candidate abnormal region into arrival position information or arrival time information of the near infrared detection position, triggering the near infrared spectrum device to conduct fixed-point sampling in a corresponding triggering time window, and conducting triggering correction on imaging delay, communication delay and sampling response delay.
7. The thermal imaging and near infrared based rubber mass sorting method of claim 6, wherein acquiring near infrared absorbance data at least one moisture characteristic wavelength comprises: Collecting near infrared spectrum data in a plurality of moisture sensitive wave bands respectively, preprocessing the near infrared spectrum data, extracting absorbance characteristics, derivative characteristics and peak shape characteristics from the preprocessed near infrared spectrum data, and collecting the absorbance characteristics, derivative characteristics and peak shape characteristics to obtain a near infrared characteristic set.
8. The method for sorting rubber blocks based on thermal imaging and near infrared according to claim 7, wherein when temperature compensation is performed on the near infrared absorbance data to obtain a temperature corrected near infrared characteristic parameter, comprising: the method comprises the steps of obtaining the surface temperature of a candidate abnormal region corresponding to the fixed point sampling, determining formula type information of a rubber block, selecting a temperature compensation relation corresponding to the formula type information according to the formula type information, calling the selected temperature compensation relation to carry out temperature correction on the near infrared characteristic set, and obtaining the temperature correction near infrared characteristic parameters.
9. The method for sorting rubber blocks based on thermal imaging and near infrared according to claim 8, wherein the sorting judgment is performed based on the temperature corrected near infrared characteristic parameter, and when the sorting judgment result is outputted, the method comprises: The method comprises the steps of correcting near infrared characteristic parameters according to temperature, calculating sorting judgment values according to the temperature correction near infrared characteristic parameters and a preset judgment rule or judgment model, comparing the sorting judgment values with sorting judgment thresholds and outputting abnormal sorting judgment results or qualified sorting judgment results, calculating confidence coefficient of the sorting judgment results and comparing the confidence coefficient with sample inclusion thresholds, including corresponding samples into sliding sample sets when the confidence coefficient meets the sample inclusion thresholds, and updating the sorting judgment thresholds or judgment parameters based on the sliding sample sets.
10. A thermal imaging and near infrared based rubber block sorting system for implementing a thermal imaging and near infrared based rubber block sorting method according to any one of claims 1-9, characterized by comprising: The device comprises a conveying separation module, a separating station, a detecting station and a sorting station, wherein the conveying separation module is configured to sequentially convey rubber blocks to be sorted along a conveying direction to pass through the separating station, the detecting station and the sorting station; The thermal imaging acquisition processing module is configured to acquire thermal imaging images of the rubber blocks at the detection station by utilizing a thermal imaging device, perform image processing on the thermal imaging images to obtain contour information and surface temperature distribution information of each rubber block, and determine position coordinates of candidate abnormal rubber blocks and candidate abnormal areas thereof based on the surface temperature distribution information; the near infrared fixed point sampling module is configured to control a near infrared spectrum device to sample the candidate abnormal region at a near infrared detection position according to the position coordinates of the candidate abnormal region, and near infrared absorbance data at least one moisture characteristic wavelength is obtained; the temperature acquisition compensation module is configured to acquire the surface temperature of the candidate abnormal region corresponding to the fixed point sampling, take the surface temperature as compensation input, and perform temperature compensation on the near infrared absorbance data to obtain a temperature correction near infrared characteristic parameter; a sorting judgment module configured to perform sorting judgment based on the temperature correction near infrared characteristic parameter, and output a sorting judgment result; And the sorting control module is configured to generate a sorting instruction according to the sorting judgment result and the position and the conveying speed of the candidate abnormal rubber blocks in the conveying direction, and control a sorting executing mechanism to execute a splitting or rejecting action on the corresponding rubber blocks at the sorting station.

Description

Rubber block sorting method and system based on thermal imaging and near infrared Technical Field The invention relates to the technical field of rubber block sorting, in particular to a rubber block sorting method and system based on thermal imaging and near infrared. Background In the rubber processing production process, the rubber block enters the subsequent steps of packaging, storage or reprocessing after the procedures of drying and the like. In order to ensure the stability of the subsequent process, the production line generally needs to sort the discharged rubber blocks on line, and separate the rubber blocks which do not meet the predetermined quality requirement from the normal conveying path. The existing online sorting equipment generally comprises a conveying mechanism, a detecting unit, a control unit and a sorting executing mechanism, wherein the detecting unit detects passing rubber blocks and outputs judging information, and the control unit drives the sorting executing mechanism to split or reject target rubber blocks according to the judging information. However, under the continuous and high-beat conveying working condition, the existing sorting system has a remarkable problem that the synchronous control precision between detection judgment and sorting execution is insufficient. If the control unit still adopts fixed delay or an open-loop triggering mode based on nominal speed to drive the actuating mechanism, the separation action and the position of the target rubber block are not matched, so that the error removal, the missing removal or the unstable diversion is caused, and the separation reliability and the stable operation of the production line are affected. Disclosure of Invention In view of the above, the invention provides a rubber block sorting method and a system based on thermal imaging and near infrared, which aim to solve the problems of incorrect sorting or missed sorting caused by insufficient synchronous control precision between sorting judgment and sorting execution and mismatching of sorting actions and target rubber block positions in the existing rubber block online sorting process. In one aspect, the invention provides a thermal imaging and near infrared based rubber block sorting method, comprising the following steps: sequentially conveying rubber blocks to be sorted along the conveying direction to pass through a separation station, a detection station and a sorting station, wherein the separation station is used for separating adhered or stacked rubber blocks and outputting single-layer dispersed rubber blocks; Acquiring thermal imaging images of the rubber blocks at the detection station by using a thermal imaging device, performing image processing on the thermal imaging images to obtain contour information and surface temperature distribution information of each rubber block, and determining position coordinates of candidate abnormal rubber blocks and candidate abnormal areas thereof based on the surface temperature distribution information; According to the position coordinates of the candidate abnormal region, controlling a near infrared spectrum device to sample the candidate abnormal region at a near infrared detection position at a fixed point to obtain near infrared absorbance data at least one moisture characteristic wavelength; Acquiring the surface temperature of the candidate abnormal region corresponding to the fixed point sampling, taking the surface temperature as compensation input, and performing temperature compensation on the near infrared absorbance data to obtain a temperature correction near infrared characteristic parameter; Sorting judgment is carried out based on the temperature correction near infrared characteristic parameters, and sorting judgment results are output; and generating a sorting instruction according to the sorting judgment result and the position and the conveying speed of the candidate abnormal rubber blocks in the conveying direction, and controlling a sorting executing mechanism to execute a splitting or rejecting action on the corresponding rubber blocks at the sorting station. Further, when the separation station performs separation treatment on the adhered or stacked rubber blocks, the method comprises the following steps: The separation station is divided into a feeding section, a stripping section and an accelerating section which are sequentially connected, the stripping section is provided with a conveying speed different from that of the feeding section and applies stripping action to the rubber block, the stripping section is provided with vibration action to the rubber block, the tail end of the stripping section is used for detecting the lamination state of the rubber block, and parameters of the conveying speed and the vibration action are adjusted according to the lamination state. Further, when performing image processing on the thermal imaging image to obtain profile information and surface temper