Abstract: The moving target image has the characteristics of unstable feature points distribution and low SNR of image acquisition, which is not very accurate. Traditional method uses the target recognition algorithm of edge contour feature segmentation, when the image is disturbed by the irregular color noise, the peak signal to noise ratio of the image fusion is low. A moving target image recognition method based on depth pixel image fusion technology is proposed. Gathering under the complex background image of moving targets were wavelet denoising processing, for image noise reduction after purification using texture adaptive quantization density estimation method for image depth super pixel feature segmentation, image principal component edge contour feature information, the input image of moving targets were vector quantization processing, realize image matching and fusion, achieve the purpose of image recognition. The simulation results show that the algorithm is used to deal with the image fusion of moving objects, the root mean square error of the pixel feature matching is low, the peak signal to noise ratio of the output image is improved, and the image recognition accuracy is improved.