Research on Object Recognition Detection Based on Image Processing and Machine Learning

In this paper, the application of image processing and machine learning in object recognition is studied. Firstly, a calculation method based on color, size and leaf area ratio is proposed. Through image preprocessing, RGB color extraction, edge detection and other steps, the classification model is trained and predicted. Secondly, a method based on image preprocessing, two-dimensional area calculation and leaf area ratio conversion is proposed to estimate the target quality by calculating the mass distribution histogram. In object recognition, a feature extraction method based on image preprocessing, standardization and convolutional neural network is proposed to train and evaluate deep learning models and realize object recognition.