The present invention relates to the field of computer vision and machine learning technology, in particular to a mobile terminal and its object detection method and device Explanation of relevant terms: Histogram of Oriented Gradient - HOG - Gradient Direction Histogram; Support Vector Machine - SVM - Support Vector Machine; Convolutional Neural Network - CNN - Convolutional Neural Network; Deformable Part based Model - DPM - Component based deformation model; Cascade - Cascade; Image Pyramids - Image Pyramids; Feature Pyramids - Feature Pyramids; Haar - Haar Background technology: Currently, many products, including smartphones and tablets, have object detection functions, such as hand detection, face detection, pedestrian detection, etc. Taking human hand detection as an example, the main purpose of human hand detection is to detect target instances of human hands in images and videos, which can be applied to gesture recognition, human-computer interaction, etc. Object detection and recognition is a highly challenging topic in the field of computer vision. Object detection is a prerequisite for object recognition and can directly affect the success rate of recognition. As an initialization step for gesture recognition, hand detection is no exception