Multi-resolution image processing are part of this concept that has a purpose to extracting the detail information of the multi-scale input image. However, in general, to process a multiscale image there are issue that need to be solved, for an instance color difference, image matching failure, and lack of data which can be observe in a same time. In the same way, observation are work where object of interest are being analyze and being perceive. Observing a multi-resolution image with a lot of object interest would be a daunting task to do. There have been many attempts to match and stitch two or more images into one but still has a chance to get failure. In this article, the author propose a multi-resolution image observation system which are focus on three improvement field successfulness rating of image matching, utilize a classification algorithm to help perceiving object interest and user-friendly interface viewer. On pre-processing, this system take a four type magnification level of image; 10 ×, 20 ×, 50 ×, 150 ×. By investigating image matching parameter to stitching group of images, handling false positive of feature point occurred, and then finally arrange and show group of synthesized images inside web application, we believe user can observe microscopic image in a cushioned way.