Tobacco leaf image online background processing system and online background processing method

Abstract

The invention discloses a tobacco leaf image online background processing system and an online background processing method. The system comprises an image acquisition module, an image preprocessing module, an image discrimination module and an image segmentation module. The online background processing method comprises the following steps: cutting an acquired image into a background frame; judgingwhether the to-be-detected image is a tobacco leaf image or not through a perceptual hash algorithm and similarity calculation; performing edge-preserving denoising processing on an image by adoptingmean filtering, cutting the image into a plurality of regions, mapping the cut regions into a background contour curve, and finally replacing the extracted contour curve with a white, black or transparent background to obtain a background-removed tobacco leaf image. According to the online background processing system, efficient and accurate deduction of the background can be realized without human participation, the image after background matting can be automatically used as the image for next analysis, manual image storage is not needed, and a good foundation is laid for online maturity recognition.

Description

technical field [0001] The invention belongs to the technical field of tobacco leaves, and in particular relates to an online background processing system and an online background processing method for tobacco leaf images with a high degree of automation and low professional requirements. Background technique [0002] When using a mobile phone or camera to take pictures, the surrounding background of the target is often difficult to remove when taking pictures. However, when analyzing the image, the surrounding background of the target will greatly interfere with the analysis results and even cause wrong judgments; in most cases, removing Background interference can only be resolved by post-processing image processing technology; image background removal (cutout) refers to the accurate extraction of foreground objects in pictures, but in many image editing technologies, background removal is still performed offline (PC client) There are many, and it is difficult to do one-cl...

AI Technical Summary

Problems solved by technology

[0002] When using a mobile phone or camera to take pictures, the surrounding background of the target is often difficult to remove when taking pictures. However, when analyzing the image, the surrounding background of the target will greatly interfere with the analysis results and even cause wrong judgments; in most cases, removing Background interference can only be resolved by post-processing image processing technology; image background removal (cutout) refers to the accurate extraction of foreground objects in pictures, but in many image editing technologies, background removal is still performed offline (PC client) There are many, and it is difficult to do one-click online processing, and most background extraction methods rely on manual selection. Under certain complex conditions, the deduction may not be clean or even some foreground parts will be deducted. For various reasons, to a large extent It affects the rapid and accurate online identification of tobacco leaf maturity by image recognition technology

[0003] For the background removal of tobacco leaf images, the classic segmentation method has the following disadvantages, so it cannot be fully applied: 1) Threshold segmentation requires the image background and the target to have two main modes of color or grayscale, while in the tobacco leaf image , the surface texture and color of the target tobacco leaves are complex and part of the structural color is similar to the background. It is easy to misjudge the internal structure as the background by using the threshold segmentation method; The boundary contour of the heterosexual discrimination image, and the selection of the starting point of the contour is the key to the application of this method

But for the tobacco leaves in the field, the overlapping of multiple layers of tobacco leaves will seriously interfere with the background processing, which brings great difficulty to the selection of the starting point of the contour

Images