An Efficient Crystal Shadow Noise Removal Method in Crystal Crystallization Process Monitoring

Abstract

The invention discloses an efficient crystal shadow noise removal method in crystal crystallization process monitoring. The method 1) uses a serial annotation algorithm for connectivity annotation of crystal images to be processed; 2) divides crystal regions. Considering the particularity of the crystallization process, it is impossible to directly use the classic image segmentation algorithm to divide the crystal region. The present invention proposes a synthetic crystal region division algorithm, that is, first using the Graham algorithm to calculate its convex hull, and then calculating the minimum circumference The coordinates (x, y) of the upper left corner of the rectangle, the width w of the rectangle, and the height h of the rectangle are extracted from the original image to obtain all independent crystal area blocks; 3) For each crystal area block, the OTSU adaptive threshold selection algorithm is used Calculate the threshold T as the segmentation threshold, perform binarization processing, and complete the removal of shadow noise of the crystal target. This method has good shadow noise removal effect and low time complexity, and is of great significance in both academic research and industrial production practice.

Description

technical field [0001] The invention relates to the fields of crystal monitoring, process control, industrial production automation, image processing and the like, and in particular relates to a noise removal technology in crystal reaction process monitoring. Background technique [0002] The crystallization process has an extremely important impact on industrial production, especially the production process of the chemical industry. This process is widely used in the fields of special chemical industry, biopharmaceutical and agricultural chemistry for solid-liquid separation, purification and solid crystal production. [0003] In the process of crystal crystallization, its morphological characteristics are often very important attributes of crystals. Different morphological characteristics such as crystal shape and size distribution will directly affect the crystal growth and the quality of the final product. Reasonable control of the morphological characteristics during th...

AI Technical Summary

Problems solved by technology

However, this method relies heavily on the significant difference between the target to be identified and the gray level of the shadow. Due to the limitations of the image acquisition equipment, the imaging resolution of the crystal target is not the same. The gray level of a certain crystal target in the same image may be different from other The gray levels of crystal shadows are similar. If the threshold method is directly used for denoising, when the threshold value is small, the crystal target and the shadow cannot be completely separated; when the threshold value is too large, some crystals with small gray levels will be treated as Remove the shadow together, which will also cause errors in the measurement of crystal properties, such as figure 2 shown

Images