The condition of the wear particle in it can indicate the working status of the machine and show the quality of lubrication oil.Because of that ,detecting the wear particle in engine lubrication oil not only can be conscious of the quality changes of lubrication oil, but also predict the fault of the machine and make sure the machine works smoothly and safely. Today there are various methods on engine lubrication oil detecting.Some study was cited in this paper to solve the problem: the wear particle must be separated from the lubrication oil to monitor under nowadays technical condition, and it also proposed one kind of new lubrication oil detecting technology under the condition that the lubrication oil and the wear particle were not separated. During the detecting, at first, smeared the lubrication oil that had collected from the bottom of engine oil casing on the glass sheet,then collected the images of them and stored in computer by using certain magnification digital camera and image collecting card, finally processed the digital images and collected wear particle' feature parameters with the analysing software,in another word, the detecting of lubrication oil wear particle was realized.The main methods and conclusions in this paper are as follows:1. Simple glass sheet digital images of lubrication oil wear particle were obtained using instruments such as microscope, camera under unseparated condition; Images were demarcated using metallographic microscope ruler images and the diagonal line index factor was 0.8840 um/pixel.2. Processing result of the digital images using image processing toolbox of MATLAB, indicated that the contrast expansion enhancement effects was better than that of the histogram heighten; and medium filtering enhancement better than mean filtering; images after smoothing were inadvisable to be sharpened. Finally the preprocessing flow of the wear particle digital images was obtained.3. The background of the preprocessed image were splited applying Domain Value method and interfering message was removed using morphological operation The program composed with MATLAB can monitor the area,girth,equivalent diameter and form factor of the swear particle. By processing the blood cell count board image whose area was known under the same condition, the study method used in this paper enhanced the detecting accuracy of the swear particle area to 93% It was revealed that the origin of accuracy error had something to do with the parameters, choosing of image processing, split of domain value, resolution of the instrument etc.4. According to the image processing flow, this paper developed a detecting system for engine lubrication oil particle with VisualBasic6.0 and Matlab6.5.This study proposed a new method for detecting the lubrication oil particle and had positive meaning to realize changing oil according to the quality of lubrication oil.