Chocolate milk instability was monitored by time series of images. It is demonstrated how a time series of images can be summarized into a single image, named a stability barcode. The stability barcode is generated by collapsing time series frames to pixel columns and arranging the pixel columns to form the stability barcode. It is demonstrated how image filtering can be used in generation of different and more specific stability barcodes. Chocolate milk defects such as layers could be seen in barcodes generated using a median filter, curdling could be seen in barcodes generated using a standard deviation filter and marbling could be seen in barcodes made using an H-dome filter. Graphs for comparing the destabilization kinetics were deduced from the stability barcodes. The stability barcodes gave a visual summary of chocolate milk instability–the kinetic graphs a precise comparison of specific types of instability.