Hyperspectral imaging technology is a rapid and nondestructive technique, which has the characteristic of combining the image and spectra, each band will represent an image, each pixel displays a spectra. Hyperpectral image can not only obtain spectral information of the samples, but also spatial information of objects representation, having great value in many field at present. In this paper, hyperspectral imaging technology is used to visualize the distribution of petroleum hydrocarbons in soil. A sample of latosol samples with different petroleum hydrocarbons content was prepared and divided into modeling samples and prediction samples. Hyperspectral images were collected. In order to avoid the interference of image background, the mask was used to eliminate the background. After extracting the average spectral region of interest modeling in the sample, the successive projection algorithm select characteristics variables. Based on the extracted feature variables, on one hand, a MLR prediction modelis established, On the other hand, the characteristics of hyperspectral image are extracted from the prediction sample. Finally, the data of the pixel on the characteristic image is substituted into the model to obtain the content distribution of the petroleum hydrocarbon. Through the method of image processing, the different contents are given different colors to realize the visualization of the distribution of petroleum hydrocarbon content in latosol. The research results show that using hyperspectral imaging technique and image processing method can realize the visualization of the distribution of petroleum hydrocarbon content in latosol, providing the basis for petroleum hydrocarbon content identification and inversion of soil in large scale.