Noncontact canned compost maturity determining method

Abstract

The invention discloses a noncontact canned compost maturity determining method comprising the following steps: S1, extracting image data at time t; S2, preprocessing; S3, based on data obtained in S2, constructing a convolutional neural network (CNN) for extracting the feature of a compost image to obtains a 255-dimensional feature vector; S4, combining the color histogram data of an thermal image in the S1 with the feature vector output by the image feature extraction CNN in the S3 to form a real-time compost feature, and normalizing the real-time compost feature; S5 performing prediction based on a long short term memory (LSTM) network; and S6, outputting a determined result. The method for real-time detection of compost state based on temperature and appearance is accurate in results and easy to operate.

Description

Technical field [0001] The invention uses the natural image information of the compost pile and the thermal imaging image taken by the thermal imager to perform feature extraction and then classification through a deep learning neural network to determine the maturity of the compost in real time, and belongs to the field of agricultural informatics. Background technique [0002] The maturity of fertilizers is extremely important for the growth of crops. The use of immature fertilizers will not only not have a positive effect on the growth of crops, but will also attract flies and worms to lay eggs and corrode the roots of crops. The proliferation of microorganisms will also cause soil hypoxia. . The fully decomposed fertilizer will not cause adverse effects on the environment and is convenient for transportation, which is beneficial to improve soil fertility and promote plant growth. [0003] The maturity is an index that reflects the degree of stabilization of the composting proc...

AI Technical Summary

Problems solved by technology

In addition, there are some differences in the surface shape c...

Images